| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9998919736415686, | |
| "eval_steps": 500, | |
| "global_step": 3471, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0014403514457527638, | |
| "grad_norm": 2.62252908668901, | |
| "learning_rate": 2.2988505747126437e-07, | |
| "loss": 3.4281, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0028807028915055276, | |
| "grad_norm": 2.6537402724592827, | |
| "learning_rate": 5.172413793103449e-07, | |
| "loss": 3.4375, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.004321054337258291, | |
| "grad_norm": 2.6461606608506973, | |
| "learning_rate": 8.045977011494253e-07, | |
| "loss": 3.4271, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.005761405783011055, | |
| "grad_norm": 2.6482412561136113, | |
| "learning_rate": 1.0919540229885058e-06, | |
| "loss": 3.4316, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.007201757228763818, | |
| "grad_norm": 2.7090530181422676, | |
| "learning_rate": 1.3793103448275862e-06, | |
| "loss": 3.4271, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.008642108674516582, | |
| "grad_norm": 2.6957713098098757, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 3.4222, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.010082460120269346, | |
| "grad_norm": 2.823975463539696, | |
| "learning_rate": 1.9540229885057475e-06, | |
| "loss": 3.4122, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.01152281156602211, | |
| "grad_norm": 2.906829198673145, | |
| "learning_rate": 2.241379310344828e-06, | |
| "loss": 3.3972, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.012963163011774873, | |
| "grad_norm": 2.959760309292492, | |
| "learning_rate": 2.5287356321839083e-06, | |
| "loss": 3.3748, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.014403514457527637, | |
| "grad_norm": 3.0140610424144882, | |
| "learning_rate": 2.8160919540229887e-06, | |
| "loss": 3.3312, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0158438659032804, | |
| "grad_norm": 3.0958728084508227, | |
| "learning_rate": 3.103448275862069e-06, | |
| "loss": 3.2621, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.017284217349033165, | |
| "grad_norm": 3.0789510621647116, | |
| "learning_rate": 3.3908045977011496e-06, | |
| "loss": 3.1805, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.01872456879478593, | |
| "grad_norm": 3.060095215151247, | |
| "learning_rate": 3.67816091954023e-06, | |
| "loss": 3.0475, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.020164920240538693, | |
| "grad_norm": 2.862619501338678, | |
| "learning_rate": 3.96551724137931e-06, | |
| "loss": 2.8765, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.021605271686291457, | |
| "grad_norm": 2.4529701308424627, | |
| "learning_rate": 4.252873563218391e-06, | |
| "loss": 2.6438, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.02304562313204422, | |
| "grad_norm": 1.9259597661186747, | |
| "learning_rate": 4.540229885057471e-06, | |
| "loss": 2.4074, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.02448597457779698, | |
| "grad_norm": 1.1944899501060937, | |
| "learning_rate": 4.8275862068965525e-06, | |
| "loss": 2.1598, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.025926326023549745, | |
| "grad_norm": 0.7043736273275593, | |
| "learning_rate": 5.114942528735632e-06, | |
| "loss": 1.986, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.02736667746930251, | |
| "grad_norm": 0.4401877202562375, | |
| "learning_rate": 5.402298850574713e-06, | |
| "loss": 1.8752, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.028807028915055273, | |
| "grad_norm": 0.29913771411892215, | |
| "learning_rate": 5.689655172413794e-06, | |
| "loss": 1.8032, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.030247380360808037, | |
| "grad_norm": 0.2501344036852339, | |
| "learning_rate": 5.977011494252874e-06, | |
| "loss": 1.762, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.0316877318065608, | |
| "grad_norm": 0.22028235935534177, | |
| "learning_rate": 6.264367816091954e-06, | |
| "loss": 1.7284, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.03312808325231356, | |
| "grad_norm": 0.20809861418922077, | |
| "learning_rate": 6.551724137931035e-06, | |
| "loss": 1.7045, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.03456843469806633, | |
| "grad_norm": 0.20520136163663816, | |
| "learning_rate": 6.839080459770115e-06, | |
| "loss": 1.6829, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.03600878614381909, | |
| "grad_norm": 0.20174383175553343, | |
| "learning_rate": 7.126436781609196e-06, | |
| "loss": 1.6605, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.03744913758957186, | |
| "grad_norm": 0.20319553328609763, | |
| "learning_rate": 7.413793103448277e-06, | |
| "loss": 1.6379, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.03888948903532462, | |
| "grad_norm": 0.2038342591080777, | |
| "learning_rate": 7.701149425287356e-06, | |
| "loss": 1.6107, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.040329840481077385, | |
| "grad_norm": 0.20440794132298992, | |
| "learning_rate": 7.988505747126438e-06, | |
| "loss": 1.5846, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.041770191926830146, | |
| "grad_norm": 0.20590893215016112, | |
| "learning_rate": 8.275862068965518e-06, | |
| "loss": 1.5547, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.04321054337258291, | |
| "grad_norm": 0.20185248245660797, | |
| "learning_rate": 8.563218390804599e-06, | |
| "loss": 1.526, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.044650894818335674, | |
| "grad_norm": 0.1991870123255739, | |
| "learning_rate": 8.85057471264368e-06, | |
| "loss": 1.4937, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.04609124626408844, | |
| "grad_norm": 0.19465384170558617, | |
| "learning_rate": 9.13793103448276e-06, | |
| "loss": 1.4598, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.0475315977098412, | |
| "grad_norm": 0.18797574210376147, | |
| "learning_rate": 9.42528735632184e-06, | |
| "loss": 1.4343, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.04897194915559396, | |
| "grad_norm": 0.18212469780560764, | |
| "learning_rate": 9.71264367816092e-06, | |
| "loss": 1.3986, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.05041230060134673, | |
| "grad_norm": 0.1851275242281937, | |
| "learning_rate": 1e-05, | |
| "loss": 1.3663, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.05185265204709949, | |
| "grad_norm": 0.1894926644233431, | |
| "learning_rate": 1.0287356321839081e-05, | |
| "loss": 1.3365, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.05329300349285226, | |
| "grad_norm": 0.19239941503471172, | |
| "learning_rate": 1.0574712643678162e-05, | |
| "loss": 1.3029, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.05473335493860502, | |
| "grad_norm": 0.1983109078749674, | |
| "learning_rate": 1.0862068965517242e-05, | |
| "loss": 1.2663, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.056173706384357786, | |
| "grad_norm": 0.20536206577251548, | |
| "learning_rate": 1.1149425287356324e-05, | |
| "loss": 1.2238, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.057614057830110546, | |
| "grad_norm": 0.20960804903533373, | |
| "learning_rate": 1.1436781609195405e-05, | |
| "loss": 1.1841, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.059054409275863314, | |
| "grad_norm": 0.2188932735565814, | |
| "learning_rate": 1.1724137931034483e-05, | |
| "loss": 1.1332, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.060494760721616074, | |
| "grad_norm": 0.22876533797063386, | |
| "learning_rate": 1.2011494252873564e-05, | |
| "loss": 1.0851, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.061935112167368835, | |
| "grad_norm": 0.23174345011010952, | |
| "learning_rate": 1.2298850574712644e-05, | |
| "loss": 1.0293, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.0633754636131216, | |
| "grad_norm": 0.23330633292890332, | |
| "learning_rate": 1.2586206896551725e-05, | |
| "loss": 0.9716, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.06481581505887436, | |
| "grad_norm": 0.24464497890825554, | |
| "learning_rate": 1.2873563218390805e-05, | |
| "loss": 0.9035, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.06625616650462712, | |
| "grad_norm": 0.25134894132627916, | |
| "learning_rate": 1.3160919540229885e-05, | |
| "loss": 0.8296, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.0676965179503799, | |
| "grad_norm": 0.2586116096970685, | |
| "learning_rate": 1.3448275862068967e-05, | |
| "loss": 0.749, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.06913686939613266, | |
| "grad_norm": 0.25755530564734574, | |
| "learning_rate": 1.3735632183908048e-05, | |
| "loss": 0.657, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.07057722084188542, | |
| "grad_norm": 0.24571389947226407, | |
| "learning_rate": 1.4022988505747128e-05, | |
| "loss": 0.5611, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.07201757228763818, | |
| "grad_norm": 0.2161386991422811, | |
| "learning_rate": 1.4310344827586209e-05, | |
| "loss": 0.4684, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.07345792373339095, | |
| "grad_norm": 0.1845577123208276, | |
| "learning_rate": 1.459770114942529e-05, | |
| "loss": 0.3929, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.07489827517914371, | |
| "grad_norm": 0.15938611776040426, | |
| "learning_rate": 1.4885057471264368e-05, | |
| "loss": 0.3274, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.07633862662489647, | |
| "grad_norm": 0.13184840828375916, | |
| "learning_rate": 1.5172413793103448e-05, | |
| "loss": 0.273, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.07777897807064924, | |
| "grad_norm": 0.10547360872838579, | |
| "learning_rate": 1.545977011494253e-05, | |
| "loss": 0.2341, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.079219329516402, | |
| "grad_norm": 0.08378821715636167, | |
| "learning_rate": 1.574712643678161e-05, | |
| "loss": 0.2152, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.08065968096215477, | |
| "grad_norm": 0.06639227255563633, | |
| "learning_rate": 1.603448275862069e-05, | |
| "loss": 0.2028, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.08210003240790753, | |
| "grad_norm": 0.05326270333294462, | |
| "learning_rate": 1.632183908045977e-05, | |
| "loss": 0.1915, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.08354038385366029, | |
| "grad_norm": 0.045250863339317625, | |
| "learning_rate": 1.6609195402298854e-05, | |
| "loss": 0.1831, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.08498073529941305, | |
| "grad_norm": 0.03997876406980682, | |
| "learning_rate": 1.6896551724137932e-05, | |
| "loss": 0.1783, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.08642108674516583, | |
| "grad_norm": 0.03683411390731312, | |
| "learning_rate": 1.7183908045977015e-05, | |
| "loss": 0.1803, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.08786143819091859, | |
| "grad_norm": 0.033494444332579816, | |
| "learning_rate": 1.7471264367816093e-05, | |
| "loss": 0.1792, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.08930178963667135, | |
| "grad_norm": 0.027545194459475855, | |
| "learning_rate": 1.7758620689655175e-05, | |
| "loss": 0.1776, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.09074214108242411, | |
| "grad_norm": 0.025534566575995447, | |
| "learning_rate": 1.8045977011494254e-05, | |
| "loss": 0.1627, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.09218249252817688, | |
| "grad_norm": 0.02432314883206978, | |
| "learning_rate": 1.8333333333333333e-05, | |
| "loss": 0.1694, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.09362284397392964, | |
| "grad_norm": 0.02296007196296768, | |
| "learning_rate": 1.8620689655172415e-05, | |
| "loss": 0.1632, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.0950631954196824, | |
| "grad_norm": 0.02231231517863281, | |
| "learning_rate": 1.8908045977011497e-05, | |
| "loss": 0.1603, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.09650354686543516, | |
| "grad_norm": 0.020592453220906963, | |
| "learning_rate": 1.9195402298850576e-05, | |
| "loss": 0.1578, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.09794389831118792, | |
| "grad_norm": 0.020380116268697165, | |
| "learning_rate": 1.9482758620689658e-05, | |
| "loss": 0.1568, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.0993842497569407, | |
| "grad_norm": 0.02128754387207953, | |
| "learning_rate": 1.9770114942528737e-05, | |
| "loss": 0.1577, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.10082460120269346, | |
| "grad_norm": 0.020077957237521784, | |
| "learning_rate": 1.9999994940288617e-05, | |
| "loss": 0.1645, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.10226495264844622, | |
| "grad_norm": 0.02081981445796709, | |
| "learning_rate": 1.999981785092774e-05, | |
| "loss": 0.1458, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.10370530409419898, | |
| "grad_norm": 0.019642666023756323, | |
| "learning_rate": 1.9999387781117715e-05, | |
| "loss": 0.1515, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.10514565553995175, | |
| "grad_norm": 0.01993767310786855, | |
| "learning_rate": 1.9998704741738657e-05, | |
| "loss": 0.1558, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.10658600698570452, | |
| "grad_norm": 0.018085760313418967, | |
| "learning_rate": 1.9997768750070442e-05, | |
| "loss": 0.1553, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.10802635843145728, | |
| "grad_norm": 0.0203452490973548, | |
| "learning_rate": 1.9996579829792263e-05, | |
| "loss": 0.15, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.10946670987721004, | |
| "grad_norm": 0.018590604714013987, | |
| "learning_rate": 1.9995138010982028e-05, | |
| "loss": 0.1492, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.1109070613229628, | |
| "grad_norm": 0.017630128077178075, | |
| "learning_rate": 1.9993443330115592e-05, | |
| "loss": 0.1474, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.11234741276871557, | |
| "grad_norm": 0.019441688441053217, | |
| "learning_rate": 1.9991495830065857e-05, | |
| "loss": 0.1509, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.11378776421446833, | |
| "grad_norm": 0.023197073067636517, | |
| "learning_rate": 1.9989295560101656e-05, | |
| "loss": 0.1512, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.11522811566022109, | |
| "grad_norm": 0.022598358372408592, | |
| "learning_rate": 1.998684257588654e-05, | |
| "loss": 0.1457, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.11666846710597385, | |
| "grad_norm": 0.018391884570447603, | |
| "learning_rate": 1.9984136939477333e-05, | |
| "loss": 0.1487, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.11810881855172663, | |
| "grad_norm": 0.01730298666489709, | |
| "learning_rate": 1.9981178719322606e-05, | |
| "loss": 0.1408, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.11954916999747939, | |
| "grad_norm": 0.018077200725101054, | |
| "learning_rate": 1.99779679902609e-05, | |
| "loss": 0.1543, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.12098952144323215, | |
| "grad_norm": 0.018875155851426084, | |
| "learning_rate": 1.9974504833518863e-05, | |
| "loss": 0.1526, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.12242987288898491, | |
| "grad_norm": 0.01550384193559367, | |
| "learning_rate": 1.9970789336709185e-05, | |
| "loss": 0.1503, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.12387022433473767, | |
| "grad_norm": 0.01563256542028533, | |
| "learning_rate": 1.9966821593828393e-05, | |
| "loss": 0.1475, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.12531057578049043, | |
| "grad_norm": 0.015801051344428534, | |
| "learning_rate": 1.9962601705254442e-05, | |
| "loss": 0.1384, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.1267509272262432, | |
| "grad_norm": 0.018919536477648418, | |
| "learning_rate": 1.995812977774421e-05, | |
| "loss": 0.146, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.12819127867199598, | |
| "grad_norm": 0.01564902677446783, | |
| "learning_rate": 1.995340592443078e-05, | |
| "loss": 0.1446, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.12963163011774873, | |
| "grad_norm": 0.01649586526356844, | |
| "learning_rate": 1.9948430264820588e-05, | |
| "loss": 0.1443, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.1310719815635015, | |
| "grad_norm": 0.016233025873589237, | |
| "learning_rate": 1.994320292479038e-05, | |
| "loss": 0.1398, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.13251233300925425, | |
| "grad_norm": 0.018370727196108234, | |
| "learning_rate": 1.993772403658405e-05, | |
| "loss": 0.1413, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.13395268445500702, | |
| "grad_norm": 0.015413145114948405, | |
| "learning_rate": 1.9931993738809288e-05, | |
| "loss": 0.1377, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.1353930359007598, | |
| "grad_norm": 0.01579533249322328, | |
| "learning_rate": 1.9926012176434054e-05, | |
| "loss": 0.1343, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.13683338734651254, | |
| "grad_norm": 0.016623384137655177, | |
| "learning_rate": 1.991977950078295e-05, | |
| "loss": 0.1362, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.13827373879226532, | |
| "grad_norm": 0.016561476846420342, | |
| "learning_rate": 1.9913295869533345e-05, | |
| "loss": 0.1383, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.13971409023801806, | |
| "grad_norm": 0.016900232091944686, | |
| "learning_rate": 1.990656144671143e-05, | |
| "loss": 0.1448, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.14115444168377084, | |
| "grad_norm": 0.018123435569779828, | |
| "learning_rate": 1.9899576402688038e-05, | |
| "loss": 0.1364, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.1425947931295236, | |
| "grad_norm": 0.01869327376515684, | |
| "learning_rate": 1.9892340914174344e-05, | |
| "loss": 0.1457, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.14403514457527636, | |
| "grad_norm": 0.017383446282339288, | |
| "learning_rate": 1.988485516421739e-05, | |
| "loss": 0.1369, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.14403514457527636, | |
| "eval_loss": 0.13828575611114502, | |
| "eval_runtime": 863.0505, | |
| "eval_samples_per_second": 2.09, | |
| "eval_steps_per_second": 0.523, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.14547549602102913, | |
| "grad_norm": 0.01727758799136872, | |
| "learning_rate": 1.9877119342195478e-05, | |
| "loss": 0.141, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.1469158474667819, | |
| "grad_norm": 0.020724335957621343, | |
| "learning_rate": 1.986913364381333e-05, | |
| "loss": 0.1417, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.14835619891253465, | |
| "grad_norm": 0.015171059899028828, | |
| "learning_rate": 1.9860898271097194e-05, | |
| "loss": 0.1361, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.14979655035828743, | |
| "grad_norm": 0.019695421273977713, | |
| "learning_rate": 1.9852413432389685e-05, | |
| "loss": 0.1314, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.15123690180404017, | |
| "grad_norm": 0.016105641997242213, | |
| "learning_rate": 1.984367934234455e-05, | |
| "loss": 0.1355, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.15267725324979295, | |
| "grad_norm": 0.017169458650654956, | |
| "learning_rate": 1.9834696221921213e-05, | |
| "loss": 0.135, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.15411760469554572, | |
| "grad_norm": 0.016758621454569307, | |
| "learning_rate": 1.98254642983792e-05, | |
| "loss": 0.1319, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.15555795614129847, | |
| "grad_norm": 0.018163113510772863, | |
| "learning_rate": 1.9815983805272378e-05, | |
| "loss": 0.1303, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.15699830758705124, | |
| "grad_norm": 0.017990247516200045, | |
| "learning_rate": 1.980625498244306e-05, | |
| "loss": 0.139, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.158438659032804, | |
| "grad_norm": 0.015967634026842183, | |
| "learning_rate": 1.9796278076015924e-05, | |
| "loss": 0.1319, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.15987901047855677, | |
| "grad_norm": 0.014829856672342604, | |
| "learning_rate": 1.9786053338391792e-05, | |
| "loss": 0.1352, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.16131936192430954, | |
| "grad_norm": 0.0172321563882635, | |
| "learning_rate": 1.9775581028241253e-05, | |
| "loss": 0.1253, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.1627597133700623, | |
| "grad_norm": 0.018596015242561326, | |
| "learning_rate": 1.97648614104981e-05, | |
| "loss": 0.1281, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.16420006481581506, | |
| "grad_norm": 0.018018526896851798, | |
| "learning_rate": 1.9753894756352643e-05, | |
| "loss": 0.1345, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.16564041626156784, | |
| "grad_norm": 0.016819098654316897, | |
| "learning_rate": 1.9742681343244853e-05, | |
| "loss": 0.1248, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.16708076770732058, | |
| "grad_norm": 0.015776677724923773, | |
| "learning_rate": 1.9731221454857322e-05, | |
| "loss": 0.1313, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.16852111915307336, | |
| "grad_norm": 0.017615190905280357, | |
| "learning_rate": 1.9719515381108093e-05, | |
| "loss": 0.1368, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.1699614705988261, | |
| "grad_norm": 0.01586176965856958, | |
| "learning_rate": 1.970756341814335e-05, | |
| "loss": 0.1308, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.17140182204457888, | |
| "grad_norm": 0.01582181491527461, | |
| "learning_rate": 1.9695365868329895e-05, | |
| "loss": 0.1334, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.17284217349033165, | |
| "grad_norm": 0.01832860755228363, | |
| "learning_rate": 1.9682923040247513e-05, | |
| "loss": 0.1303, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.1742825249360844, | |
| "grad_norm": 0.01608043519957003, | |
| "learning_rate": 1.9670235248681154e-05, | |
| "loss": 0.128, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.17572287638183717, | |
| "grad_norm": 0.015462595330781244, | |
| "learning_rate": 1.965730281461299e-05, | |
| "loss": 0.1318, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.17716322782758992, | |
| "grad_norm": 0.01829748858884658, | |
| "learning_rate": 1.964412606521428e-05, | |
| "loss": 0.128, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.1786035792733427, | |
| "grad_norm": 0.01741437351404647, | |
| "learning_rate": 1.9630705333837096e-05, | |
| "loss": 0.129, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.18004393071909547, | |
| "grad_norm": 0.01714971316628412, | |
| "learning_rate": 1.9617040960005883e-05, | |
| "loss": 0.1274, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.18148428216484822, | |
| "grad_norm": 0.017341345727170193, | |
| "learning_rate": 1.9603133289408883e-05, | |
| "loss": 0.1285, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.182924633610601, | |
| "grad_norm": 0.018545013302290372, | |
| "learning_rate": 1.9588982673889373e-05, | |
| "loss": 0.129, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.18436498505635376, | |
| "grad_norm": 0.01917072532666851, | |
| "learning_rate": 1.9574589471436794e-05, | |
| "loss": 0.1246, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.1858053365021065, | |
| "grad_norm": 0.017610415450957716, | |
| "learning_rate": 1.955995404617765e-05, | |
| "loss": 0.1296, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.18724568794785929, | |
| "grad_norm": 0.017348619206735666, | |
| "learning_rate": 1.9545076768366336e-05, | |
| "loss": 0.1201, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.18868603939361203, | |
| "grad_norm": 0.018402248035990453, | |
| "learning_rate": 1.9529958014375748e-05, | |
| "loss": 0.1288, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.1901263908393648, | |
| "grad_norm": 0.017347671383931933, | |
| "learning_rate": 1.9514598166687772e-05, | |
| "loss": 0.1322, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.19156674228511758, | |
| "grad_norm": 0.01862965351110765, | |
| "learning_rate": 1.9498997613883597e-05, | |
| "loss": 0.1368, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.19300709373087033, | |
| "grad_norm": 0.016601710581881732, | |
| "learning_rate": 1.9483156750633906e-05, | |
| "loss": 0.1326, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.1944474451766231, | |
| "grad_norm": 0.015529988935684115, | |
| "learning_rate": 1.946707597768886e-05, | |
| "loss": 0.1224, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.19588779662237585, | |
| "grad_norm": 0.02017712606978146, | |
| "learning_rate": 1.9450755701867994e-05, | |
| "loss": 0.1291, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.19732814806812862, | |
| "grad_norm": 0.0167232245419185, | |
| "learning_rate": 1.9434196336049897e-05, | |
| "loss": 0.1201, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.1987684995138814, | |
| "grad_norm": 0.019048290037720794, | |
| "learning_rate": 1.941739829916177e-05, | |
| "loss": 0.1241, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.20020885095963414, | |
| "grad_norm": 0.017899758339100985, | |
| "learning_rate": 1.940036201616886e-05, | |
| "loss": 0.1255, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.20164920240538692, | |
| "grad_norm": 0.018753604988563825, | |
| "learning_rate": 1.9383087918063662e-05, | |
| "loss": 0.1279, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.20308955385113966, | |
| "grad_norm": 0.016720002206463015, | |
| "learning_rate": 1.9365576441855046e-05, | |
| "loss": 0.1265, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.20452990529689244, | |
| "grad_norm": 0.01598156321133028, | |
| "learning_rate": 1.9347828030557196e-05, | |
| "loss": 0.1247, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.2059702567426452, | |
| "grad_norm": 0.015176452766976297, | |
| "learning_rate": 1.932984313317839e-05, | |
| "loss": 0.1269, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.20741060818839796, | |
| "grad_norm": 0.01743363415449858, | |
| "learning_rate": 1.931162220470967e-05, | |
| "loss": 0.1252, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.20885095963415073, | |
| "grad_norm": 0.020263642348868235, | |
| "learning_rate": 1.9293165706113287e-05, | |
| "loss": 0.1245, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.2102913110799035, | |
| "grad_norm": 0.016973378857152475, | |
| "learning_rate": 1.9274474104311083e-05, | |
| "loss": 0.1229, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.21173166252565626, | |
| "grad_norm": 0.01631083673087625, | |
| "learning_rate": 1.925554787217265e-05, | |
| "loss": 0.124, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.21317201397140903, | |
| "grad_norm": 0.019965918558652533, | |
| "learning_rate": 1.9236387488503378e-05, | |
| "loss": 0.1223, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.21461236541716178, | |
| "grad_norm": 0.021848215070733194, | |
| "learning_rate": 1.921699343803235e-05, | |
| "loss": 0.1138, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.21605271686291455, | |
| "grad_norm": 0.018291368502582516, | |
| "learning_rate": 1.9197366211400058e-05, | |
| "loss": 0.1241, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.21749306830866733, | |
| "grad_norm": 0.019059330643677385, | |
| "learning_rate": 1.9177506305146008e-05, | |
| "loss": 0.1268, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.21893341975442007, | |
| "grad_norm": 0.017080958515049458, | |
| "learning_rate": 1.9157414221696155e-05, | |
| "loss": 0.1206, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.22037377120017285, | |
| "grad_norm": 0.017957542038359017, | |
| "learning_rate": 1.9137090469350185e-05, | |
| "loss": 0.1244, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.2218141226459256, | |
| "grad_norm": 0.01802711782041451, | |
| "learning_rate": 1.9116535562268658e-05, | |
| "loss": 0.1206, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.22325447409167837, | |
| "grad_norm": 0.019287045682764212, | |
| "learning_rate": 1.9095750020460017e-05, | |
| "loss": 0.1144, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.22469482553743114, | |
| "grad_norm": 0.01879207545472608, | |
| "learning_rate": 1.9074734369767407e-05, | |
| "loss": 0.1192, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.2261351769831839, | |
| "grad_norm": 0.019461601988728348, | |
| "learning_rate": 1.9053489141855386e-05, | |
| "loss": 0.125, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.22757552842893666, | |
| "grad_norm": 0.019720624818386844, | |
| "learning_rate": 1.9032014874196476e-05, | |
| "loss": 0.1269, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.22901587987468944, | |
| "grad_norm": 0.018318990062697504, | |
| "learning_rate": 1.901031211005756e-05, | |
| "loss": 0.1203, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.23045623132044218, | |
| "grad_norm": 0.019164348842944955, | |
| "learning_rate": 1.898838139848614e-05, | |
| "loss": 0.1213, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.23189658276619496, | |
| "grad_norm": 0.02213208810579334, | |
| "learning_rate": 1.8966223294296445e-05, | |
| "loss": 0.1213, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.2333369342119477, | |
| "grad_norm": 0.018200798899224192, | |
| "learning_rate": 1.8943838358055403e-05, | |
| "loss": 0.1275, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.23477728565770048, | |
| "grad_norm": 0.018399660059020002, | |
| "learning_rate": 1.892122715606846e-05, | |
| "loss": 0.1226, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.23621763710345325, | |
| "grad_norm": 0.017674503935350343, | |
| "learning_rate": 1.8898390260365227e-05, | |
| "loss": 0.1175, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.237657988549206, | |
| "grad_norm": 0.019168528020990075, | |
| "learning_rate": 1.8875328248685047e-05, | |
| "loss": 0.1228, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.23909833999495878, | |
| "grad_norm": 0.01707510254798075, | |
| "learning_rate": 1.885204170446235e-05, | |
| "loss": 0.1229, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.24053869144071152, | |
| "grad_norm": 0.017143155499374434, | |
| "learning_rate": 1.8828531216811912e-05, | |
| "loss": 0.1163, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.2419790428864643, | |
| "grad_norm": 0.01972260397119316, | |
| "learning_rate": 1.8804797380513944e-05, | |
| "loss": 0.1151, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.24341939433221707, | |
| "grad_norm": 0.017300491914188608, | |
| "learning_rate": 1.878084079599903e-05, | |
| "loss": 0.1197, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.24485974577796982, | |
| "grad_norm": 0.017415521885546983, | |
| "learning_rate": 1.8756662069332966e-05, | |
| "loss": 0.12, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.2463000972237226, | |
| "grad_norm": 0.020650688836768367, | |
| "learning_rate": 1.8732261812201408e-05, | |
| "loss": 0.1159, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.24774044866947534, | |
| "grad_norm": 0.022394976870943134, | |
| "learning_rate": 1.8707640641894395e-05, | |
| "loss": 0.1182, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.2491808001152281, | |
| "grad_norm": 0.018481000136883082, | |
| "learning_rate": 1.8682799181290747e-05, | |
| "loss": 0.117, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.25062115156098086, | |
| "grad_norm": 0.020812551732477143, | |
| "learning_rate": 1.86577380588423e-05, | |
| "loss": 0.1174, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.25206150300673363, | |
| "grad_norm": 0.019146509121457756, | |
| "learning_rate": 1.8632457908558006e-05, | |
| "loss": 0.1128, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.2535018544524864, | |
| "grad_norm": 0.02088872645363683, | |
| "learning_rate": 1.8606959369987885e-05, | |
| "loss": 0.1258, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.2549422058982392, | |
| "grad_norm": 0.020975712319762363, | |
| "learning_rate": 1.8581243088206865e-05, | |
| "loss": 0.1252, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.25638255734399196, | |
| "grad_norm": 0.022104093266461965, | |
| "learning_rate": 1.8555309713798445e-05, | |
| "loss": 0.1175, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.2578229087897447, | |
| "grad_norm": 0.01935436681312094, | |
| "learning_rate": 1.8529159902838253e-05, | |
| "loss": 0.1155, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.25926326023549745, | |
| "grad_norm": 0.02103943800048682, | |
| "learning_rate": 1.8502794316877423e-05, | |
| "loss": 0.1163, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.2607036116812502, | |
| "grad_norm": 0.02064951760573597, | |
| "learning_rate": 1.8476213622925885e-05, | |
| "loss": 0.1151, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.262143963127003, | |
| "grad_norm": 0.02391509020619219, | |
| "learning_rate": 1.844941849343548e-05, | |
| "loss": 0.1154, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.2635843145727558, | |
| "grad_norm": 0.019299067245624735, | |
| "learning_rate": 1.842240960628294e-05, | |
| "loss": 0.118, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.2650246660185085, | |
| "grad_norm": 0.018999660132613904, | |
| "learning_rate": 1.8395187644752756e-05, | |
| "loss": 0.1121, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.26646501746426127, | |
| "grad_norm": 0.019465104345890567, | |
| "learning_rate": 1.8367753297519874e-05, | |
| "loss": 0.1147, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.26790536891001404, | |
| "grad_norm": 0.017798540374291294, | |
| "learning_rate": 1.8340107258632288e-05, | |
| "loss": 0.1192, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.2693457203557668, | |
| "grad_norm": 0.022579684079520896, | |
| "learning_rate": 1.831225022749347e-05, | |
| "loss": 0.1127, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.2707860718015196, | |
| "grad_norm": 0.019409928418770574, | |
| "learning_rate": 1.828418290884468e-05, | |
| "loss": 0.1141, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.2722264232472723, | |
| "grad_norm": 0.019982776222918414, | |
| "learning_rate": 1.8255906012747137e-05, | |
| "loss": 0.1107, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.2736667746930251, | |
| "grad_norm": 0.019327451944760602, | |
| "learning_rate": 1.8227420254564066e-05, | |
| "loss": 0.1088, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.27510712613877786, | |
| "grad_norm": 0.02195579382510436, | |
| "learning_rate": 1.819872635494258e-05, | |
| "loss": 0.1149, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.27654747758453063, | |
| "grad_norm": 0.019331091074997867, | |
| "learning_rate": 1.816982503979546e-05, | |
| "loss": 0.1187, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.2779878290302834, | |
| "grad_norm": 0.0198267215148203, | |
| "learning_rate": 1.8140717040282797e-05, | |
| "loss": 0.1133, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.2794281804760361, | |
| "grad_norm": 0.021536265466576106, | |
| "learning_rate": 1.811140309279348e-05, | |
| "loss": 0.1169, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.2808685319217889, | |
| "grad_norm": 0.02522795945903232, | |
| "learning_rate": 1.808188393892658e-05, | |
| "loss": 0.1162, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.2823088833675417, | |
| "grad_norm": 0.021202903232191053, | |
| "learning_rate": 1.805216032547258e-05, | |
| "loss": 0.1177, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.28374923481329445, | |
| "grad_norm": 0.02708284619788456, | |
| "learning_rate": 1.8022233004394487e-05, | |
| "loss": 0.1189, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.2851895862590472, | |
| "grad_norm": 0.02891719428802073, | |
| "learning_rate": 1.7992102732808798e-05, | |
| "loss": 0.116, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.2866299377048, | |
| "grad_norm": 0.020017133174120256, | |
| "learning_rate": 1.796177027296637e-05, | |
| "loss": 0.1192, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.2880702891505527, | |
| "grad_norm": 0.022814840523695007, | |
| "learning_rate": 1.79312363922331e-05, | |
| "loss": 0.1196, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2880702891505527, | |
| "eval_loss": 0.1148042306303978, | |
| "eval_runtime": 862.8924, | |
| "eval_samples_per_second": 2.091, | |
| "eval_steps_per_second": 0.523, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2895106405963055, | |
| "grad_norm": 0.020165318251700003, | |
| "learning_rate": 1.7900501863070552e-05, | |
| "loss": 0.1092, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.29095099204205827, | |
| "grad_norm": 0.020276491773740017, | |
| "learning_rate": 1.7869567463016394e-05, | |
| "loss": 0.108, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.29239134348781104, | |
| "grad_norm": 0.025430907092320074, | |
| "learning_rate": 1.7838433974664714e-05, | |
| "loss": 0.1198, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.2938316949335638, | |
| "grad_norm": 0.02600654612802859, | |
| "learning_rate": 1.7807102185646247e-05, | |
| "loss": 0.1164, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.29527204637931653, | |
| "grad_norm": 0.0193830589897909, | |
| "learning_rate": 1.7775572888608438e-05, | |
| "loss": 0.1151, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.2967123978250693, | |
| "grad_norm": 0.023599658745454725, | |
| "learning_rate": 1.774384688119539e-05, | |
| "loss": 0.1183, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.2981527492708221, | |
| "grad_norm": 0.021674931585504164, | |
| "learning_rate": 1.7711924966027678e-05, | |
| "loss": 0.1141, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.29959310071657486, | |
| "grad_norm": 0.023299214871684407, | |
| "learning_rate": 1.767980795068206e-05, | |
| "loss": 0.1194, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.30103345216232763, | |
| "grad_norm": 0.02147089523180919, | |
| "learning_rate": 1.7647496647671033e-05, | |
| "loss": 0.1123, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.30247380360808035, | |
| "grad_norm": 0.021457716083701384, | |
| "learning_rate": 1.761499187442228e-05, | |
| "loss": 0.1058, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.3039141550538331, | |
| "grad_norm": 0.01969637917151117, | |
| "learning_rate": 1.7582294453257996e-05, | |
| "loss": 0.1207, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.3053545064995859, | |
| "grad_norm": 0.023839310938145133, | |
| "learning_rate": 1.7549405211374072e-05, | |
| "loss": 0.1146, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.3067948579453387, | |
| "grad_norm": 0.01832092911473367, | |
| "learning_rate": 1.7516324980819185e-05, | |
| "loss": 0.113, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.30823520939109145, | |
| "grad_norm": 0.02124767953755396, | |
| "learning_rate": 1.7483054598473734e-05, | |
| "loss": 0.1213, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.30967556083684417, | |
| "grad_norm": 0.0247043392947563, | |
| "learning_rate": 1.7449594906028684e-05, | |
| "loss": 0.1121, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.31111591228259694, | |
| "grad_norm": 0.022139995366462776, | |
| "learning_rate": 1.7415946749964252e-05, | |
| "loss": 0.1132, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.3125562637283497, | |
| "grad_norm": 0.01979540959946202, | |
| "learning_rate": 1.7382110981528506e-05, | |
| "loss": 0.1085, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.3139966151741025, | |
| "grad_norm": 0.029924025152490485, | |
| "learning_rate": 1.734808845671583e-05, | |
| "loss": 0.1054, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.31543696661985526, | |
| "grad_norm": 0.023093772502185278, | |
| "learning_rate": 1.7313880036245257e-05, | |
| "loss": 0.112, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.316877318065608, | |
| "grad_norm": 0.023473379108296808, | |
| "learning_rate": 1.7279486585538712e-05, | |
| "loss": 0.1135, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.31831766951136076, | |
| "grad_norm": 0.021733393242624945, | |
| "learning_rate": 1.7244908974699112e-05, | |
| "loss": 0.1134, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.31975802095711353, | |
| "grad_norm": 0.021271929042974584, | |
| "learning_rate": 1.721014807848833e-05, | |
| "loss": 0.1164, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.3211983724028663, | |
| "grad_norm": 0.020487120222693576, | |
| "learning_rate": 1.7175204776305102e-05, | |
| "loss": 0.1077, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.3226387238486191, | |
| "grad_norm": 0.021171582329437753, | |
| "learning_rate": 1.7140079952162765e-05, | |
| "loss": 0.1102, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.32407907529437185, | |
| "grad_norm": 0.02144446773648067, | |
| "learning_rate": 1.7104774494666877e-05, | |
| "loss": 0.1092, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.3255194267401246, | |
| "grad_norm": 0.019040039144316633, | |
| "learning_rate": 1.7069289296992756e-05, | |
| "loss": 0.112, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.32695977818587735, | |
| "grad_norm": 0.021286149625859865, | |
| "learning_rate": 1.703362525686288e-05, | |
| "loss": 0.1106, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 0.3284001296316301, | |
| "grad_norm": 0.02183865564006019, | |
| "learning_rate": 1.6997783276524177e-05, | |
| "loss": 0.1134, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.3298404810773829, | |
| "grad_norm": 0.021387717982638638, | |
| "learning_rate": 1.6961764262725187e-05, | |
| "loss": 0.1063, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 0.33128083252313567, | |
| "grad_norm": 0.024865173996299898, | |
| "learning_rate": 1.6925569126693135e-05, | |
| "loss": 0.1051, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.3327211839688884, | |
| "grad_norm": 0.023180712867703127, | |
| "learning_rate": 1.6889198784110883e-05, | |
| "loss": 0.1058, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 0.33416153541464116, | |
| "grad_norm": 0.02218441265430226, | |
| "learning_rate": 1.6852654155093745e-05, | |
| "loss": 0.1117, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.33560188686039394, | |
| "grad_norm": 0.02165110006319973, | |
| "learning_rate": 1.681593616416623e-05, | |
| "loss": 0.112, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 0.3370422383061467, | |
| "grad_norm": 0.02200923978026833, | |
| "learning_rate": 1.6779045740238643e-05, | |
| "loss": 0.114, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.3384825897518995, | |
| "grad_norm": 0.02197865224880007, | |
| "learning_rate": 1.6741983816583583e-05, | |
| "loss": 0.111, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.3399229411976522, | |
| "grad_norm": 0.019554896311713962, | |
| "learning_rate": 1.6704751330812342e-05, | |
| "loss": 0.1019, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.341363292643405, | |
| "grad_norm": 0.024134291900428952, | |
| "learning_rate": 1.666734922485117e-05, | |
| "loss": 0.1092, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 0.34280364408915776, | |
| "grad_norm": 0.022673698274783088, | |
| "learning_rate": 1.662977844491746e-05, | |
| "loss": 0.113, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.34424399553491053, | |
| "grad_norm": 0.02603003971238878, | |
| "learning_rate": 1.6592039941495803e-05, | |
| "loss": 0.1036, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 0.3456843469806633, | |
| "grad_norm": 0.024554078349019966, | |
| "learning_rate": 1.6554134669313943e-05, | |
| "loss": 0.1121, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.347124698426416, | |
| "grad_norm": 0.019717425570643314, | |
| "learning_rate": 1.6516063587318627e-05, | |
| "loss": 0.1068, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 0.3485650498721688, | |
| "grad_norm": 0.027429511589759405, | |
| "learning_rate": 1.647782765865134e-05, | |
| "loss": 0.106, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.35000540131792157, | |
| "grad_norm": 0.026628925564535798, | |
| "learning_rate": 1.6439427850623944e-05, | |
| "loss": 0.1123, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 0.35144575276367435, | |
| "grad_norm": 0.026461597229239776, | |
| "learning_rate": 1.64008651346942e-05, | |
| "loss": 0.1099, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.3528861042094271, | |
| "grad_norm": 0.02584266985928258, | |
| "learning_rate": 1.63621404864412e-05, | |
| "loss": 0.1107, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.35432645565517984, | |
| "grad_norm": 0.019646883384690197, | |
| "learning_rate": 1.6323254885540672e-05, | |
| "loss": 0.1121, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.3557668071009326, | |
| "grad_norm": 0.024026012804653142, | |
| "learning_rate": 1.6284209315740225e-05, | |
| "loss": 0.1047, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 0.3572071585466854, | |
| "grad_norm": 0.024476248259603612, | |
| "learning_rate": 1.6245004764834423e-05, | |
| "loss": 0.109, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.35864750999243816, | |
| "grad_norm": 0.022810087697062553, | |
| "learning_rate": 1.620564222463982e-05, | |
| "loss": 0.1102, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 0.36008786143819094, | |
| "grad_norm": 0.024569475912673062, | |
| "learning_rate": 1.6166122690969872e-05, | |
| "loss": 0.1057, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.36152821288394366, | |
| "grad_norm": 0.026160188797535698, | |
| "learning_rate": 1.612644716360972e-05, | |
| "loss": 0.0995, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 0.36296856432969643, | |
| "grad_norm": 0.025602704231673833, | |
| "learning_rate": 1.6086616646290926e-05, | |
| "loss": 0.116, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.3644089157754492, | |
| "grad_norm": 0.02307850757189033, | |
| "learning_rate": 1.6046632146666056e-05, | |
| "loss": 0.1088, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 0.365849267221202, | |
| "grad_norm": 0.024139060873310914, | |
| "learning_rate": 1.60064946762832e-05, | |
| "loss": 0.1044, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.36728961866695475, | |
| "grad_norm": 0.023315249645455345, | |
| "learning_rate": 1.5966205250560393e-05, | |
| "loss": 0.1119, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.36872997011270753, | |
| "grad_norm": 0.021641266926506458, | |
| "learning_rate": 1.592576488875989e-05, | |
| "loss": 0.1094, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.37017032155846025, | |
| "grad_norm": 0.022047478963055464, | |
| "learning_rate": 1.5885174613962427e-05, | |
| "loss": 0.1083, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 0.371610673004213, | |
| "grad_norm": 0.02400330121786872, | |
| "learning_rate": 1.5844435453041294e-05, | |
| "loss": 0.1139, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.3730510244499658, | |
| "grad_norm": 0.022595131555882185, | |
| "learning_rate": 1.5803548436636394e-05, | |
| "loss": 0.108, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 0.37449137589571857, | |
| "grad_norm": 0.02791419433017582, | |
| "learning_rate": 1.576251459912814e-05, | |
| "loss": 0.1134, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.37593172734147134, | |
| "grad_norm": 0.026002680251400723, | |
| "learning_rate": 1.5721334978611307e-05, | |
| "loss": 0.1107, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 0.37737207878722406, | |
| "grad_norm": 0.022484735806660563, | |
| "learning_rate": 1.5680010616868762e-05, | |
| "loss": 0.1076, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.37881243023297684, | |
| "grad_norm": 0.021483070318961833, | |
| "learning_rate": 1.5638542559345106e-05, | |
| "loss": 0.1034, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 0.3802527816787296, | |
| "grad_norm": 0.023871643967031505, | |
| "learning_rate": 1.559693185512023e-05, | |
| "loss": 0.1106, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.3816931331244824, | |
| "grad_norm": 0.021906873584555547, | |
| "learning_rate": 1.555517955688277e-05, | |
| "loss": 0.1077, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.38313348457023516, | |
| "grad_norm": 0.023737665896899265, | |
| "learning_rate": 1.5513286720903488e-05, | |
| "loss": 0.1153, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.3845738360159879, | |
| "grad_norm": 0.022433920258610178, | |
| "learning_rate": 1.5471254407008526e-05, | |
| "loss": 0.1037, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 0.38601418746174065, | |
| "grad_norm": 0.02261886005036298, | |
| "learning_rate": 1.542908367855263e-05, | |
| "loss": 0.1088, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.38745453890749343, | |
| "grad_norm": 0.021421057577652544, | |
| "learning_rate": 1.53867756023922e-05, | |
| "loss": 0.1047, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 0.3888948903532462, | |
| "grad_norm": 0.021709509569311863, | |
| "learning_rate": 1.534433124885836e-05, | |
| "loss": 0.1033, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.390335241798999, | |
| "grad_norm": 0.024820443979506786, | |
| "learning_rate": 1.530175169172982e-05, | |
| "loss": 0.1097, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 0.3917755932447517, | |
| "grad_norm": 0.025227643673449818, | |
| "learning_rate": 1.525903800820575e-05, | |
| "loss": 0.1065, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.39321594469050447, | |
| "grad_norm": 0.022487840207955263, | |
| "learning_rate": 1.5216191278878522e-05, | |
| "loss": 0.1064, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 0.39465629613625725, | |
| "grad_norm": 0.025854963249344314, | |
| "learning_rate": 1.517321258770636e-05, | |
| "loss": 0.1043, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.39609664758201, | |
| "grad_norm": 0.02260216955018553, | |
| "learning_rate": 1.5130103021985929e-05, | |
| "loss": 0.1126, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.3975369990277628, | |
| "grad_norm": 0.02491639306364899, | |
| "learning_rate": 1.5086863672324826e-05, | |
| "loss": 0.1039, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.3989773504735155, | |
| "grad_norm": 0.022899995574611955, | |
| "learning_rate": 1.5043495632613982e-05, | |
| "loss": 0.1066, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 0.4004177019192683, | |
| "grad_norm": 0.024707905868095445, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.1106, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.40185805336502106, | |
| "grad_norm": 0.0241361491731915, | |
| "learning_rate": 1.4956377874857395e-05, | |
| "loss": 0.1001, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 0.40329840481077384, | |
| "grad_norm": 0.025677892301256368, | |
| "learning_rate": 1.4912630360760743e-05, | |
| "loss": 0.1092, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.4047387562565266, | |
| "grad_norm": 0.024135758692475445, | |
| "learning_rate": 1.4868758564456785e-05, | |
| "loss": 0.1116, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 0.40617910770227933, | |
| "grad_norm": 0.024527918000747655, | |
| "learning_rate": 1.4824763595836404e-05, | |
| "loss": 0.1065, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.4076194591480321, | |
| "grad_norm": 0.023548606886532042, | |
| "learning_rate": 1.4780646567906571e-05, | |
| "loss": 0.1009, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 0.4090598105937849, | |
| "grad_norm": 0.026633678739013535, | |
| "learning_rate": 1.473640859676217e-05, | |
| "loss": 0.1059, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.41050016203953765, | |
| "grad_norm": 0.02593995796499118, | |
| "learning_rate": 1.4692050801557769e-05, | |
| "loss": 0.107, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.4119405134852904, | |
| "grad_norm": 0.02413535724648029, | |
| "learning_rate": 1.4647574304479295e-05, | |
| "loss": 0.1115, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.4133808649310432, | |
| "grad_norm": 0.02297878729642286, | |
| "learning_rate": 1.4602980230715674e-05, | |
| "loss": 0.1083, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 0.4148212163767959, | |
| "grad_norm": 0.02697991820571532, | |
| "learning_rate": 1.4558269708430333e-05, | |
| "loss": 0.1051, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.4162615678225487, | |
| "grad_norm": 0.025004735182102, | |
| "learning_rate": 1.4513443868732674e-05, | |
| "loss": 0.1103, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 0.41770191926830147, | |
| "grad_norm": 0.024578324660308907, | |
| "learning_rate": 1.4468503845649462e-05, | |
| "loss": 0.108, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.41914227071405424, | |
| "grad_norm": 0.024290400744172484, | |
| "learning_rate": 1.4423450776096122e-05, | |
| "loss": 0.102, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 0.420582622159807, | |
| "grad_norm": 0.025652851866900635, | |
| "learning_rate": 1.4378285799848004e-05, | |
| "loss": 0.1139, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.42202297360555974, | |
| "grad_norm": 0.022542605080051747, | |
| "learning_rate": 1.4333010059511505e-05, | |
| "loss": 0.1048, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 0.4234633250513125, | |
| "grad_norm": 0.022564970591101585, | |
| "learning_rate": 1.4287624700495211e-05, | |
| "loss": 0.1053, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.4249036764970653, | |
| "grad_norm": 0.024332336625839772, | |
| "learning_rate": 1.4242130870980882e-05, | |
| "loss": 0.1035, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.42634402794281806, | |
| "grad_norm": 0.02565641763751157, | |
| "learning_rate": 1.4196529721894427e-05, | |
| "loss": 0.1054, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.42778437938857083, | |
| "grad_norm": 0.022329221067199423, | |
| "learning_rate": 1.4150822406876774e-05, | |
| "loss": 0.1064, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 0.42922473083432355, | |
| "grad_norm": 0.02525079350860865, | |
| "learning_rate": 1.4105010082254697e-05, | |
| "loss": 0.1089, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.43066508228007633, | |
| "grad_norm": 0.025470140748940266, | |
| "learning_rate": 1.4059093907011552e-05, | |
| "loss": 0.103, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 0.4321054337258291, | |
| "grad_norm": 0.03889636426311705, | |
| "learning_rate": 1.401307504275796e-05, | |
| "loss": 0.106, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.4321054337258291, | |
| "eval_loss": 0.10669519007205963, | |
| "eval_runtime": 863.9882, | |
| "eval_samples_per_second": 2.088, | |
| "eval_steps_per_second": 0.522, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.4335457851715819, | |
| "grad_norm": 0.03107255388021069, | |
| "learning_rate": 1.3966954653702423e-05, | |
| "loss": 0.1086, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 0.43498613661733465, | |
| "grad_norm": 0.029782191799490893, | |
| "learning_rate": 1.3920733906621861e-05, | |
| "loss": 0.1056, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.43642648806308737, | |
| "grad_norm": 0.03596510921750815, | |
| "learning_rate": 1.3874413970832123e-05, | |
| "loss": 0.1057, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 0.43786683950884014, | |
| "grad_norm": 0.02690673392192122, | |
| "learning_rate": 1.3827996018158356e-05, | |
| "loss": 0.1009, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.4393071909545929, | |
| "grad_norm": 0.02415574342794587, | |
| "learning_rate": 1.378148122290541e-05, | |
| "loss": 0.1092, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 0.4407475424003457, | |
| "grad_norm": 0.025267363786179952, | |
| "learning_rate": 1.3734870761828095e-05, | |
| "loss": 0.1046, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.44218789384609847, | |
| "grad_norm": 0.02251899862464666, | |
| "learning_rate": 1.368816581410143e-05, | |
| "loss": 0.1026, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 0.4436282452918512, | |
| "grad_norm": 0.024093194788389363, | |
| "learning_rate": 1.3641367561290795e-05, | |
| "loss": 0.1069, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.44506859673760396, | |
| "grad_norm": 0.02865301350630449, | |
| "learning_rate": 1.3594477187322065e-05, | |
| "loss": 0.101, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 0.44650894818335674, | |
| "grad_norm": 0.025991749936729126, | |
| "learning_rate": 1.3547495878451635e-05, | |
| "loss": 0.104, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.4479492996291095, | |
| "grad_norm": 0.02427741302197225, | |
| "learning_rate": 1.3500424823236413e-05, | |
| "loss": 0.1097, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 0.4493896510748623, | |
| "grad_norm": 0.025286951386541058, | |
| "learning_rate": 1.3453265212503756e-05, | |
| "loss": 0.1025, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.450830002520615, | |
| "grad_norm": 0.02293738313311065, | |
| "learning_rate": 1.340601823932135e-05, | |
| "loss": 0.1026, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 0.4522703539663678, | |
| "grad_norm": 0.029909633188798766, | |
| "learning_rate": 1.335868509896702e-05, | |
| "loss": 0.1093, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.45371070541212055, | |
| "grad_norm": 0.02489864832302057, | |
| "learning_rate": 1.3311266988898477e-05, | |
| "loss": 0.1075, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 0.4551510568578733, | |
| "grad_norm": 0.02626703651775361, | |
| "learning_rate": 1.3263765108723061e-05, | |
| "loss": 0.1052, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.4565914083036261, | |
| "grad_norm": 0.028232314920455838, | |
| "learning_rate": 1.3216180660167355e-05, | |
| "loss": 0.103, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 0.4580317597493789, | |
| "grad_norm": 0.02893621210041688, | |
| "learning_rate": 1.3168514847046802e-05, | |
| "loss": 0.102, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.4594721111951316, | |
| "grad_norm": 0.027792315343772488, | |
| "learning_rate": 1.3120768875235252e-05, | |
| "loss": 0.1006, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 0.46091246264088437, | |
| "grad_norm": 0.022941453299660446, | |
| "learning_rate": 1.3072943952634446e-05, | |
| "loss": 0.101, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.46235281408663714, | |
| "grad_norm": 0.02754317441994994, | |
| "learning_rate": 1.3025041289143459e-05, | |
| "loss": 0.1014, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 0.4637931655323899, | |
| "grad_norm": 0.023409812236038566, | |
| "learning_rate": 1.2977062096628096e-05, | |
| "loss": 0.1018, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.4652335169781427, | |
| "grad_norm": 0.02736910879882447, | |
| "learning_rate": 1.2929007588890241e-05, | |
| "loss": 0.1029, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 0.4666738684238954, | |
| "grad_norm": 0.024599830487748515, | |
| "learning_rate": 1.2880878981637129e-05, | |
| "loss": 0.1076, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.4681142198696482, | |
| "grad_norm": 0.030038012520127225, | |
| "learning_rate": 1.2832677492450602e-05, | |
| "loss": 0.1058, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 0.46955457131540096, | |
| "grad_norm": 0.027368913953633472, | |
| "learning_rate": 1.2784404340756315e-05, | |
| "loss": 0.1082, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.47099492276115373, | |
| "grad_norm": 0.02334017117279336, | |
| "learning_rate": 1.2736060747792877e-05, | |
| "loss": 0.106, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 0.4724352742069065, | |
| "grad_norm": 0.02870373822056335, | |
| "learning_rate": 1.268764793658094e-05, | |
| "loss": 0.1099, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.4738756256526592, | |
| "grad_norm": 0.029235939335636004, | |
| "learning_rate": 1.2639167131892294e-05, | |
| "loss": 0.0973, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 0.475315977098412, | |
| "grad_norm": 0.026146387485585586, | |
| "learning_rate": 1.2590619560218851e-05, | |
| "loss": 0.1087, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.4767563285441648, | |
| "grad_norm": 0.02842049987719084, | |
| "learning_rate": 1.2542006449741631e-05, | |
| "loss": 0.1061, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 0.47819667998991755, | |
| "grad_norm": 0.024518713018732985, | |
| "learning_rate": 1.249332903029969e-05, | |
| "loss": 0.1051, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.4796370314356703, | |
| "grad_norm": 0.02484058398034788, | |
| "learning_rate": 1.2444588533358996e-05, | |
| "loss": 0.1022, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 0.48107738288142304, | |
| "grad_norm": 0.03724000521562738, | |
| "learning_rate": 1.23957861919813e-05, | |
| "loss": 0.1056, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.4825177343271758, | |
| "grad_norm": 0.03293802155567722, | |
| "learning_rate": 1.2346923240792907e-05, | |
| "loss": 0.1092, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 0.4839580857729286, | |
| "grad_norm": 0.026650789981673747, | |
| "learning_rate": 1.229800091595347e-05, | |
| "loss": 0.1008, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.48539843721868137, | |
| "grad_norm": 0.02758805567603314, | |
| "learning_rate": 1.2249020455124703e-05, | |
| "loss": 0.1024, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 0.48683878866443414, | |
| "grad_norm": 0.0289561047779867, | |
| "learning_rate": 1.2199983097439079e-05, | |
| "loss": 0.0974, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.48827914011018686, | |
| "grad_norm": 0.02472150617570116, | |
| "learning_rate": 1.2150890083468465e-05, | |
| "loss": 0.1017, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 0.48971949155593963, | |
| "grad_norm": 0.026933463536791147, | |
| "learning_rate": 1.2101742655192761e-05, | |
| "loss": 0.107, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.4911598430016924, | |
| "grad_norm": 0.02704841529205261, | |
| "learning_rate": 1.2052542055968461e-05, | |
| "loss": 0.1049, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 0.4926001944474452, | |
| "grad_norm": 0.029435211264724145, | |
| "learning_rate": 1.2003289530497206e-05, | |
| "loss": 0.1097, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.49404054589319796, | |
| "grad_norm": 0.023704099909174187, | |
| "learning_rate": 1.1953986324794295e-05, | |
| "loss": 0.1027, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 0.4954808973389507, | |
| "grad_norm": 0.024368103569431525, | |
| "learning_rate": 1.1904633686157158e-05, | |
| "loss": 0.1078, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.49692124878470345, | |
| "grad_norm": 0.024017154477919515, | |
| "learning_rate": 1.1855232863133809e-05, | |
| "loss": 0.1009, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 0.4983616002304562, | |
| "grad_norm": 0.03151983841636764, | |
| "learning_rate": 1.1805785105491247e-05, | |
| "loss": 0.1021, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.499801951676209, | |
| "grad_norm": 0.02991694116471432, | |
| "learning_rate": 1.1756291664183858e-05, | |
| "loss": 0.1076, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 0.5012423031219617, | |
| "grad_norm": 0.03657521191046876, | |
| "learning_rate": 1.1706753791321748e-05, | |
| "loss": 0.1042, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.5026826545677145, | |
| "grad_norm": 0.0279482378393294, | |
| "learning_rate": 1.1657172740139074e-05, | |
| "loss": 0.1017, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 0.5041230060134673, | |
| "grad_norm": 0.030069339039746814, | |
| "learning_rate": 1.1607549764962342e-05, | |
| "loss": 0.1042, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.5055633574592201, | |
| "grad_norm": 0.026487495544772206, | |
| "learning_rate": 1.1557886121178683e-05, | |
| "loss": 0.1079, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 0.5070037089049728, | |
| "grad_norm": 0.024984858540099013, | |
| "learning_rate": 1.1508183065204066e-05, | |
| "loss": 0.1037, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.5084440603507255, | |
| "grad_norm": 0.026914213670237396, | |
| "learning_rate": 1.1458441854451539e-05, | |
| "loss": 0.1001, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 0.5098844117964784, | |
| "grad_norm": 0.029179893358272806, | |
| "learning_rate": 1.1408663747299409e-05, | |
| "loss": 0.1002, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.5113247632422311, | |
| "grad_norm": 0.039586221088331666, | |
| "learning_rate": 1.13588500030594e-05, | |
| "loss": 0.105, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 0.5127651146879839, | |
| "grad_norm": 0.03249368292701708, | |
| "learning_rate": 1.130900188194481e-05, | |
| "loss": 0.1031, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.5142054661337366, | |
| "grad_norm": 0.024055776588308493, | |
| "learning_rate": 1.1259120645038612e-05, | |
| "loss": 0.1013, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 0.5156458175794894, | |
| "grad_norm": 0.025863370886123525, | |
| "learning_rate": 1.1209207554261573e-05, | |
| "loss": 0.1, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.5170861690252422, | |
| "grad_norm": 0.025434390514411754, | |
| "learning_rate": 1.1159263872340293e-05, | |
| "loss": 0.1007, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 0.5185265204709949, | |
| "grad_norm": 0.02745842348721705, | |
| "learning_rate": 1.1109290862775307e-05, | |
| "loss": 0.096, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.5199668719167477, | |
| "grad_norm": 0.027615229169377083, | |
| "learning_rate": 1.1059289789809071e-05, | |
| "loss": 0.0998, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 0.5214072233625004, | |
| "grad_norm": 0.027453704008570706, | |
| "learning_rate": 1.1009261918394028e-05, | |
| "loss": 0.0972, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.5228475748082532, | |
| "grad_norm": 0.024700516688583, | |
| "learning_rate": 1.0959208514160561e-05, | |
| "loss": 0.1055, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 0.524287926254006, | |
| "grad_norm": 0.032125782290431566, | |
| "learning_rate": 1.0909130843385009e-05, | |
| "loss": 0.0979, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.5257282776997587, | |
| "grad_norm": 0.027487991421855863, | |
| "learning_rate": 1.085903017295761e-05, | |
| "loss": 0.0992, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 0.5271686291455115, | |
| "grad_norm": 0.0284479042668781, | |
| "learning_rate": 1.0808907770350463e-05, | |
| "loss": 0.098, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.5286089805912643, | |
| "grad_norm": 0.02721705482015998, | |
| "learning_rate": 1.0758764903585457e-05, | |
| "loss": 0.1025, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 0.530049332037017, | |
| "grad_norm": 0.028561724576000452, | |
| "learning_rate": 1.070860284120219e-05, | |
| "loss": 0.1002, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.5314896834827698, | |
| "grad_norm": 0.030215649153775434, | |
| "learning_rate": 1.0658422852225889e-05, | |
| "loss": 0.113, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 0.5329300349285225, | |
| "grad_norm": 0.023003505373751384, | |
| "learning_rate": 1.0608226206135292e-05, | |
| "loss": 0.1004, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.5343703863742754, | |
| "grad_norm": 0.032778870546567584, | |
| "learning_rate": 1.0558014172830537e-05, | |
| "loss": 0.1026, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 0.5358107378200281, | |
| "grad_norm": 0.024228471157562194, | |
| "learning_rate": 1.0507788022601033e-05, | |
| "loss": 0.0959, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.5372510892657808, | |
| "grad_norm": 0.027347316934608, | |
| "learning_rate": 1.0457549026093338e-05, | |
| "loss": 0.0994, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 0.5386914407115336, | |
| "grad_norm": 0.02492119460153467, | |
| "learning_rate": 1.0407298454278983e-05, | |
| "loss": 0.096, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.5401317921572864, | |
| "grad_norm": 0.02730713294802567, | |
| "learning_rate": 1.0357037578422349e-05, | |
| "loss": 0.1063, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 0.5415721436030392, | |
| "grad_norm": 0.02753505593775209, | |
| "learning_rate": 1.0306767670048497e-05, | |
| "loss": 0.1004, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.5430124950487919, | |
| "grad_norm": 0.027125750017208197, | |
| "learning_rate": 1.0256490000910986e-05, | |
| "loss": 0.1033, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 0.5444528464945446, | |
| "grad_norm": 0.02526923620994891, | |
| "learning_rate": 1.0206205842959718e-05, | |
| "loss": 0.1044, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.5458931979402974, | |
| "grad_norm": 0.03456611614472051, | |
| "learning_rate": 1.0155916468308749e-05, | |
| "loss": 0.1018, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 0.5473335493860502, | |
| "grad_norm": 0.02687752666688773, | |
| "learning_rate": 1.0105623149204118e-05, | |
| "loss": 0.1071, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.548773900831803, | |
| "grad_norm": 0.0348842935948145, | |
| "learning_rate": 1.0055327157991639e-05, | |
| "loss": 0.1018, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 0.5502142522775557, | |
| "grad_norm": 0.03047489251408279, | |
| "learning_rate": 1.0005029767084739e-05, | |
| "loss": 0.1029, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.5516546037233084, | |
| "grad_norm": 0.03295680008205151, | |
| "learning_rate": 9.954732248932243e-06, | |
| "loss": 0.1042, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 0.5530949551690613, | |
| "grad_norm": 0.026523017503203378, | |
| "learning_rate": 9.904435875986213e-06, | |
| "loss": 0.0989, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.554535306614814, | |
| "grad_norm": 0.03199960655757483, | |
| "learning_rate": 9.85414192066972e-06, | |
| "loss": 0.1034, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 0.5559756580605668, | |
| "grad_norm": 0.03321351438487483, | |
| "learning_rate": 9.803851655344682e-06, | |
| "loss": 0.0954, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.5574160095063195, | |
| "grad_norm": 0.026379907025225486, | |
| "learning_rate": 9.75356635227966e-06, | |
| "loss": 0.1007, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 0.5588563609520723, | |
| "grad_norm": 0.03215809081400103, | |
| "learning_rate": 9.70328728361769e-06, | |
| "loss": 0.1041, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.5602967123978251, | |
| "grad_norm": 0.028935560901012115, | |
| "learning_rate": 9.653015721344073e-06, | |
| "loss": 0.104, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 0.5617370638435778, | |
| "grad_norm": 0.026479371672269946, | |
| "learning_rate": 9.602752937254215e-06, | |
| "loss": 0.0987, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.5631774152893306, | |
| "grad_norm": 0.02735022790208615, | |
| "learning_rate": 9.552500202921449e-06, | |
| "loss": 0.1052, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 0.5646177667350833, | |
| "grad_norm": 0.029074239132167123, | |
| "learning_rate": 9.502258789664865e-06, | |
| "loss": 0.1069, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.5660581181808362, | |
| "grad_norm": 0.03497134937258824, | |
| "learning_rate": 9.45202996851714e-06, | |
| "loss": 0.1068, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 0.5674984696265889, | |
| "grad_norm": 0.03121000622963146, | |
| "learning_rate": 9.401815010192388e-06, | |
| "loss": 0.1058, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.5689388210723416, | |
| "grad_norm": 0.029762366581440217, | |
| "learning_rate": 9.351615185054029e-06, | |
| "loss": 0.1027, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 0.5703791725180944, | |
| "grad_norm": 0.031933869149627876, | |
| "learning_rate": 9.301431763082623e-06, | |
| "loss": 0.1001, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.5718195239638472, | |
| "grad_norm": 0.024595673857129345, | |
| "learning_rate": 9.251266013843757e-06, | |
| "loss": 0.0993, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 0.5732598754096, | |
| "grad_norm": 0.02931011040644607, | |
| "learning_rate": 9.201119206455927e-06, | |
| "loss": 0.107, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.5747002268553527, | |
| "grad_norm": 0.02841549880839795, | |
| "learning_rate": 9.150992609558425e-06, | |
| "loss": 0.1013, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 0.5761405783011054, | |
| "grad_norm": 0.029699203984976344, | |
| "learning_rate": 9.10088749127926e-06, | |
| "loss": 0.1003, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.5761405783011054, | |
| "eval_loss": 0.10270440578460693, | |
| "eval_runtime": 862.9139, | |
| "eval_samples_per_second": 2.091, | |
| "eval_steps_per_second": 0.523, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.5775809297468583, | |
| "grad_norm": 0.02555642653773717, | |
| "learning_rate": 9.050805119203035e-06, | |
| "loss": 0.0964, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 0.579021281192611, | |
| "grad_norm": 0.03206482511392409, | |
| "learning_rate": 9.000746760338938e-06, | |
| "loss": 0.1061, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.5804616326383638, | |
| "grad_norm": 0.029335358877190958, | |
| "learning_rate": 8.950713681088647e-06, | |
| "loss": 0.0989, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 0.5819019840841165, | |
| "grad_norm": 0.029061894519663043, | |
| "learning_rate": 8.900707147214301e-06, | |
| "loss": 0.1041, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.5833423355298692, | |
| "grad_norm": 0.027789281243274542, | |
| "learning_rate": 8.850728423806487e-06, | |
| "loss": 0.1002, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 0.5847826869756221, | |
| "grad_norm": 0.028006016395305748, | |
| "learning_rate": 8.800778775252221e-06, | |
| "loss": 0.0971, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.5862230384213748, | |
| "grad_norm": 0.029135758984445807, | |
| "learning_rate": 8.75085946520298e-06, | |
| "loss": 0.1089, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 0.5876633898671276, | |
| "grad_norm": 0.028264974680774237, | |
| "learning_rate": 8.700971756542707e-06, | |
| "loss": 0.0999, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.5891037413128803, | |
| "grad_norm": 0.02787134319938327, | |
| "learning_rate": 8.65111691135589e-06, | |
| "loss": 0.0996, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 0.5905440927586331, | |
| "grad_norm": 0.02930265638543135, | |
| "learning_rate": 8.601296190895611e-06, | |
| "loss": 0.0998, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.5919844442043859, | |
| "grad_norm": 0.028136073425125765, | |
| "learning_rate": 8.551510855551658e-06, | |
| "loss": 0.1022, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 0.5934247956501386, | |
| "grad_norm": 0.030075075847505933, | |
| "learning_rate": 8.501762164818615e-06, | |
| "loss": 0.1011, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.5948651470958914, | |
| "grad_norm": 0.02509232015436124, | |
| "learning_rate": 8.452051377264025e-06, | |
| "loss": 0.1089, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 0.5963054985416442, | |
| "grad_norm": 0.029019522531841563, | |
| "learning_rate": 8.402379750496535e-06, | |
| "loss": 0.1043, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.5977458499873969, | |
| "grad_norm": 0.02946442999816807, | |
| "learning_rate": 8.35274854113407e-06, | |
| "loss": 0.1033, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 0.5991862014331497, | |
| "grad_norm": 0.028470170352907883, | |
| "learning_rate": 8.303159004772065e-06, | |
| "loss": 0.1015, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.6006265528789024, | |
| "grad_norm": 0.027074007980222393, | |
| "learning_rate": 8.253612395951697e-06, | |
| "loss": 0.1077, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 0.6020669043246553, | |
| "grad_norm": 0.02594322523858361, | |
| "learning_rate": 8.204109968128128e-06, | |
| "loss": 0.1052, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.603507255770408, | |
| "grad_norm": 0.027981527626155637, | |
| "learning_rate": 8.15465297363881e-06, | |
| "loss": 0.1076, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 0.6049476072161607, | |
| "grad_norm": 0.027128594062002694, | |
| "learning_rate": 8.105242663671807e-06, | |
| "loss": 0.1019, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.6063879586619135, | |
| "grad_norm": 0.02399321575005835, | |
| "learning_rate": 8.055880288234127e-06, | |
| "loss": 0.1014, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 0.6078283101076662, | |
| "grad_norm": 0.030384147510662384, | |
| "learning_rate": 8.006567096120103e-06, | |
| "loss": 0.102, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.6092686615534191, | |
| "grad_norm": 0.0275729769794035, | |
| "learning_rate": 7.957304334879809e-06, | |
| "loss": 0.1066, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 0.6107090129991718, | |
| "grad_norm": 0.02827555110704183, | |
| "learning_rate": 7.908093250787496e-06, | |
| "loss": 0.1026, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.6121493644449245, | |
| "grad_norm": 0.0280447976215247, | |
| "learning_rate": 7.858935088810054e-06, | |
| "loss": 0.1041, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 0.6135897158906773, | |
| "grad_norm": 0.026871658750667743, | |
| "learning_rate": 7.809831092575528e-06, | |
| "loss": 0.1033, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.6150300673364301, | |
| "grad_norm": 0.029232871601405933, | |
| "learning_rate": 7.760782504341644e-06, | |
| "loss": 0.1047, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 0.6164704187821829, | |
| "grad_norm": 0.023805191017251096, | |
| "learning_rate": 7.7117905649644e-06, | |
| "loss": 0.0983, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.6179107702279356, | |
| "grad_norm": 0.03103222984517563, | |
| "learning_rate": 7.662856513866643e-06, | |
| "loss": 0.0968, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 0.6193511216736883, | |
| "grad_norm": 0.028105627160877313, | |
| "learning_rate": 7.613981589006754e-06, | |
| "loss": 0.099, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.6207914731194412, | |
| "grad_norm": 0.030597646841460515, | |
| "learning_rate": 7.565167026847294e-06, | |
| "loss": 0.0979, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 0.6222318245651939, | |
| "grad_norm": 0.02970594057212837, | |
| "learning_rate": 7.5164140623237454e-06, | |
| "loss": 0.0975, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.6236721760109467, | |
| "grad_norm": 0.027159849019462532, | |
| "learning_rate": 7.467723928813262e-06, | |
| "loss": 0.1036, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 0.6251125274566994, | |
| "grad_norm": 0.03257508113397351, | |
| "learning_rate": 7.419097858103464e-06, | |
| "loss": 0.0997, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.6265528789024521, | |
| "grad_norm": 0.029509877961538443, | |
| "learning_rate": 7.370537080361282e-06, | |
| "loss": 0.0977, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 0.627993230348205, | |
| "grad_norm": 0.02914606675872259, | |
| "learning_rate": 7.3220428241018225e-06, | |
| "loss": 0.0961, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.6294335817939577, | |
| "grad_norm": 0.028245193180560037, | |
| "learning_rate": 7.273616316157312e-06, | |
| "loss": 0.1009, | |
| "step": 2185 | |
| }, | |
| { | |
| "epoch": 0.6308739332397105, | |
| "grad_norm": 0.031258732018464554, | |
| "learning_rate": 7.225258781646036e-06, | |
| "loss": 0.0959, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.6323142846854632, | |
| "grad_norm": 0.029187877030262547, | |
| "learning_rate": 7.176971443941359e-06, | |
| "loss": 0.0984, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 0.633754636131216, | |
| "grad_norm": 0.027481848533676127, | |
| "learning_rate": 7.128755524640771e-06, | |
| "loss": 0.1023, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.6351949875769688, | |
| "grad_norm": 0.028467126169710644, | |
| "learning_rate": 7.080612243534981e-06, | |
| "loss": 0.104, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 0.6366353390227215, | |
| "grad_norm": 0.02584742929400372, | |
| "learning_rate": 7.032542818577066e-06, | |
| "loss": 0.0968, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.6380756904684743, | |
| "grad_norm": 0.024188950488983806, | |
| "learning_rate": 6.984548465851641e-06, | |
| "loss": 0.0965, | |
| "step": 2215 | |
| }, | |
| { | |
| "epoch": 0.6395160419142271, | |
| "grad_norm": 0.028851010852631174, | |
| "learning_rate": 6.93663039954412e-06, | |
| "loss": 0.1007, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.6409563933599798, | |
| "grad_norm": 0.025328516360271506, | |
| "learning_rate": 6.888789831909972e-06, | |
| "loss": 0.101, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 0.6423967448057326, | |
| "grad_norm": 0.03102519024419765, | |
| "learning_rate": 6.841027973244077e-06, | |
| "loss": 0.098, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.6438370962514853, | |
| "grad_norm": 0.037884633433375385, | |
| "learning_rate": 6.793346031850085e-06, | |
| "loss": 0.0987, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 0.6452774476972382, | |
| "grad_norm": 0.03111478820563393, | |
| "learning_rate": 6.745745214009866e-06, | |
| "loss": 0.0984, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.6467177991429909, | |
| "grad_norm": 0.02761748015741369, | |
| "learning_rate": 6.698226723952985e-06, | |
| "loss": 0.0996, | |
| "step": 2245 | |
| }, | |
| { | |
| "epoch": 0.6481581505887437, | |
| "grad_norm": 0.02546847357291813, | |
| "learning_rate": 6.65079176382623e-06, | |
| "loss": 0.1027, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.6495985020344964, | |
| "grad_norm": 0.028225946201781264, | |
| "learning_rate": 6.603441533663214e-06, | |
| "loss": 0.1052, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 0.6510388534802491, | |
| "grad_norm": 0.028818402757150797, | |
| "learning_rate": 6.556177231354003e-06, | |
| "loss": 0.103, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.652479204926002, | |
| "grad_norm": 0.027109318220465493, | |
| "learning_rate": 6.509000052614824e-06, | |
| "loss": 0.1043, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 0.6539195563717547, | |
| "grad_norm": 0.02806157217123479, | |
| "learning_rate": 6.4619111909577995e-06, | |
| "loss": 0.0962, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.6553599078175075, | |
| "grad_norm": 0.029350220106511105, | |
| "learning_rate": 6.414911837660768e-06, | |
| "loss": 0.0997, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 0.6568002592632602, | |
| "grad_norm": 0.02863157441645804, | |
| "learning_rate": 6.36800318173714e-06, | |
| "loss": 0.1004, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.658240610709013, | |
| "grad_norm": 0.030549045202254512, | |
| "learning_rate": 6.321186409905812e-06, | |
| "loss": 0.0995, | |
| "step": 2285 | |
| }, | |
| { | |
| "epoch": 0.6596809621547658, | |
| "grad_norm": 0.03217236024450049, | |
| "learning_rate": 6.274462706561153e-06, | |
| "loss": 0.1008, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.6611213136005185, | |
| "grad_norm": 0.02822548256388095, | |
| "learning_rate": 6.227833253743045e-06, | |
| "loss": 0.0962, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 0.6625616650462713, | |
| "grad_norm": 0.02614875322287693, | |
| "learning_rate": 6.181299231106963e-06, | |
| "loss": 0.0987, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.6640020164920241, | |
| "grad_norm": 0.02541722852556565, | |
| "learning_rate": 6.134861815894146e-06, | |
| "loss": 0.0967, | |
| "step": 2305 | |
| }, | |
| { | |
| "epoch": 0.6654423679377768, | |
| "grad_norm": 0.02850719972847212, | |
| "learning_rate": 6.08852218290181e-06, | |
| "loss": 0.0947, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.6668827193835296, | |
| "grad_norm": 0.02908443110544134, | |
| "learning_rate": 6.0422815044534265e-06, | |
| "loss": 0.1037, | |
| "step": 2315 | |
| }, | |
| { | |
| "epoch": 0.6683230708292823, | |
| "grad_norm": 0.029315617072749215, | |
| "learning_rate": 5.9961409503690605e-06, | |
| "loss": 0.0963, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.6697634222750352, | |
| "grad_norm": 0.030033042623501118, | |
| "learning_rate": 5.950101687935783e-06, | |
| "loss": 0.1027, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 0.6712037737207879, | |
| "grad_norm": 0.02915046992535852, | |
| "learning_rate": 5.904164881878143e-06, | |
| "loss": 0.1055, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.6726441251665406, | |
| "grad_norm": 0.025236471457550445, | |
| "learning_rate": 5.858331694328686e-06, | |
| "loss": 0.0982, | |
| "step": 2335 | |
| }, | |
| { | |
| "epoch": 0.6740844766122934, | |
| "grad_norm": 0.03401543932639446, | |
| "learning_rate": 5.812603284798575e-06, | |
| "loss": 0.0949, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.6755248280580461, | |
| "grad_norm": 0.028026329462367786, | |
| "learning_rate": 5.766980810148238e-06, | |
| "loss": 0.1, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 0.676965179503799, | |
| "grad_norm": 0.02874201116713264, | |
| "learning_rate": 5.721465424558111e-06, | |
| "loss": 0.0958, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.6784055309495517, | |
| "grad_norm": 0.028963974093940446, | |
| "learning_rate": 5.676058279499438e-06, | |
| "loss": 0.0989, | |
| "step": 2355 | |
| }, | |
| { | |
| "epoch": 0.6798458823953044, | |
| "grad_norm": 0.029496429086982617, | |
| "learning_rate": 5.630760523705137e-06, | |
| "loss": 0.0969, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.6812862338410572, | |
| "grad_norm": 0.02472939959975296, | |
| "learning_rate": 5.585573303140741e-06, | |
| "loss": 0.1005, | |
| "step": 2365 | |
| }, | |
| { | |
| "epoch": 0.68272658528681, | |
| "grad_norm": 0.02943173144839556, | |
| "learning_rate": 5.540497760975412e-06, | |
| "loss": 0.0991, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.6841669367325628, | |
| "grad_norm": 0.027039554476503823, | |
| "learning_rate": 5.495535037553003e-06, | |
| "loss": 0.0967, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 0.6856072881783155, | |
| "grad_norm": 0.03130069538494267, | |
| "learning_rate": 5.450686270363244e-06, | |
| "loss": 0.104, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.6870476396240682, | |
| "grad_norm": 0.02792170269632637, | |
| "learning_rate": 5.405952594012921e-06, | |
| "loss": 0.0971, | |
| "step": 2385 | |
| }, | |
| { | |
| "epoch": 0.6884879910698211, | |
| "grad_norm": 0.026536938565135147, | |
| "learning_rate": 5.361335140197199e-06, | |
| "loss": 0.0998, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.6899283425155738, | |
| "grad_norm": 0.030495085484657028, | |
| "learning_rate": 5.316835037670987e-06, | |
| "loss": 0.0989, | |
| "step": 2395 | |
| }, | |
| { | |
| "epoch": 0.6913686939613266, | |
| "grad_norm": 0.025636540838806603, | |
| "learning_rate": 5.272453412220389e-06, | |
| "loss": 0.1005, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.6928090454070793, | |
| "grad_norm": 0.029276688852897748, | |
| "learning_rate": 5.228191386634212e-06, | |
| "loss": 0.096, | |
| "step": 2405 | |
| }, | |
| { | |
| "epoch": 0.694249396852832, | |
| "grad_norm": 0.028892951411287924, | |
| "learning_rate": 5.184050080675558e-06, | |
| "loss": 0.0982, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.6956897482985849, | |
| "grad_norm": 0.029668168507339023, | |
| "learning_rate": 5.140030611053512e-06, | |
| "loss": 0.0983, | |
| "step": 2415 | |
| }, | |
| { | |
| "epoch": 0.6971300997443376, | |
| "grad_norm": 0.029331223786389957, | |
| "learning_rate": 5.096134091394879e-06, | |
| "loss": 0.1039, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.6985704511900904, | |
| "grad_norm": 0.028007436697092276, | |
| "learning_rate": 5.052361632216009e-06, | |
| "loss": 0.0952, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 0.7000108026358431, | |
| "grad_norm": 0.028924268558975137, | |
| "learning_rate": 5.008714340894716e-06, | |
| "loss": 0.1025, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.7014511540815959, | |
| "grad_norm": 0.028546541140812298, | |
| "learning_rate": 4.965193321642245e-06, | |
| "loss": 0.1051, | |
| "step": 2435 | |
| }, | |
| { | |
| "epoch": 0.7028915055273487, | |
| "grad_norm": 0.03228047586525536, | |
| "learning_rate": 4.9217996754753505e-06, | |
| "loss": 0.1042, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.7043318569731014, | |
| "grad_norm": 0.026889053511112666, | |
| "learning_rate": 4.878534500188443e-06, | |
| "loss": 0.0979, | |
| "step": 2445 | |
| }, | |
| { | |
| "epoch": 0.7057722084188542, | |
| "grad_norm": 0.024493931252593228, | |
| "learning_rate": 4.835398890325806e-06, | |
| "loss": 0.1006, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.707212559864607, | |
| "grad_norm": 0.031072455568760707, | |
| "learning_rate": 4.792393937153914e-06, | |
| "loss": 0.1, | |
| "step": 2455 | |
| }, | |
| { | |
| "epoch": 0.7086529113103597, | |
| "grad_norm": 0.03526863560562569, | |
| "learning_rate": 4.749520728633812e-06, | |
| "loss": 0.1004, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.7100932627561125, | |
| "grad_norm": 0.027375718409826295, | |
| "learning_rate": 4.706780349393621e-06, | |
| "loss": 0.099, | |
| "step": 2465 | |
| }, | |
| { | |
| "epoch": 0.7115336142018652, | |
| "grad_norm": 0.029579970429738103, | |
| "learning_rate": 4.664173880701065e-06, | |
| "loss": 0.1008, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.7129739656476181, | |
| "grad_norm": 0.031758604921792485, | |
| "learning_rate": 4.62170240043614e-06, | |
| "loss": 0.0983, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 0.7144143170933708, | |
| "grad_norm": 0.02656899768489825, | |
| "learning_rate": 4.579366983063829e-06, | |
| "loss": 0.0997, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.7158546685391235, | |
| "grad_norm": 0.02566962597338034, | |
| "learning_rate": 4.537168699606928e-06, | |
| "loss": 0.1015, | |
| "step": 2485 | |
| }, | |
| { | |
| "epoch": 0.7172950199848763, | |
| "grad_norm": 0.028715798458422183, | |
| "learning_rate": 4.4951086176189666e-06, | |
| "loss": 0.107, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.718735371430629, | |
| "grad_norm": 0.02887905888346922, | |
| "learning_rate": 4.453187801157155e-06, | |
| "loss": 0.0984, | |
| "step": 2495 | |
| }, | |
| { | |
| "epoch": 0.7201757228763819, | |
| "grad_norm": 0.028332689752910086, | |
| "learning_rate": 4.411407310755513e-06, | |
| "loss": 0.1007, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.7201757228763819, | |
| "eval_loss": 0.10072407871484756, | |
| "eval_runtime": 862.0549, | |
| "eval_samples_per_second": 2.093, | |
| "eval_steps_per_second": 0.523, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.7216160743221346, | |
| "grad_norm": 0.02833390362499488, | |
| "learning_rate": 4.369768203398014e-06, | |
| "loss": 0.0957, | |
| "step": 2505 | |
| }, | |
| { | |
| "epoch": 0.7230564257678873, | |
| "grad_norm": 0.030945876414709347, | |
| "learning_rate": 4.328271532491859e-06, | |
| "loss": 0.102, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 0.7244967772136401, | |
| "grad_norm": 0.029283752981204014, | |
| "learning_rate": 4.2869183478408125e-06, | |
| "loss": 0.0985, | |
| "step": 2515 | |
| }, | |
| { | |
| "epoch": 0.7259371286593929, | |
| "grad_norm": 0.027428758917667755, | |
| "learning_rate": 4.2457096956186525e-06, | |
| "loss": 0.1018, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.7273774801051457, | |
| "grad_norm": 0.03315591743445466, | |
| "learning_rate": 4.2046466183427035e-06, | |
| "loss": 0.1075, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 0.7288178315508984, | |
| "grad_norm": 0.030025762956048904, | |
| "learning_rate": 4.163730154847462e-06, | |
| "loss": 0.0929, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 0.7302581829966511, | |
| "grad_norm": 0.030103449099341675, | |
| "learning_rate": 4.122961340258312e-06, | |
| "loss": 0.0964, | |
| "step": 2535 | |
| }, | |
| { | |
| "epoch": 0.731698534442404, | |
| "grad_norm": 0.02917754573138957, | |
| "learning_rate": 4.082341205965344e-06, | |
| "loss": 0.0993, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.7331388858881567, | |
| "grad_norm": 0.02789369472409532, | |
| "learning_rate": 4.0418707795972575e-06, | |
| "loss": 0.0995, | |
| "step": 2545 | |
| }, | |
| { | |
| "epoch": 0.7345792373339095, | |
| "grad_norm": 0.029979889727084676, | |
| "learning_rate": 4.001551084995363e-06, | |
| "loss": 0.1031, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.7360195887796622, | |
| "grad_norm": 0.027974195684437115, | |
| "learning_rate": 3.961383142187691e-06, | |
| "loss": 0.1021, | |
| "step": 2555 | |
| }, | |
| { | |
| "epoch": 0.7374599402254151, | |
| "grad_norm": 0.030070715679730322, | |
| "learning_rate": 3.9213679673631745e-06, | |
| "loss": 0.1029, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.7389002916711678, | |
| "grad_norm": 0.027066646976217957, | |
| "learning_rate": 3.881506572845933e-06, | |
| "loss": 0.1056, | |
| "step": 2565 | |
| }, | |
| { | |
| "epoch": 0.7403406431169205, | |
| "grad_norm": 0.0284274448607563, | |
| "learning_rate": 3.841799967069686e-06, | |
| "loss": 0.1075, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 0.7417809945626733, | |
| "grad_norm": 0.028074887598320608, | |
| "learning_rate": 3.8022491545522346e-06, | |
| "loss": 0.103, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 0.743221346008426, | |
| "grad_norm": 0.03333775657686949, | |
| "learning_rate": 3.7628551358700303e-06, | |
| "loss": 0.0997, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.7446616974541789, | |
| "grad_norm": 0.027376231042512, | |
| "learning_rate": 3.723618907632882e-06, | |
| "loss": 0.1024, | |
| "step": 2585 | |
| }, | |
| { | |
| "epoch": 0.7461020488999316, | |
| "grad_norm": 0.031319210954457215, | |
| "learning_rate": 3.6845414624587326e-06, | |
| "loss": 0.1032, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 0.7475424003456843, | |
| "grad_norm": 0.028908867069655585, | |
| "learning_rate": 3.6456237889485547e-06, | |
| "loss": 0.1023, | |
| "step": 2595 | |
| }, | |
| { | |
| "epoch": 0.7489827517914371, | |
| "grad_norm": 0.027194027978343076, | |
| "learning_rate": 3.606866871661333e-06, | |
| "loss": 0.0976, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.7504231032371899, | |
| "grad_norm": 0.028443079367809734, | |
| "learning_rate": 3.5682716910891613e-06, | |
| "loss": 0.0986, | |
| "step": 2605 | |
| }, | |
| { | |
| "epoch": 0.7518634546829427, | |
| "grad_norm": 0.030731013519078837, | |
| "learning_rate": 3.5298392236324365e-06, | |
| "loss": 0.1009, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.7533038061286954, | |
| "grad_norm": 0.0285250306739096, | |
| "learning_rate": 3.491570441575154e-06, | |
| "loss": 0.0989, | |
| "step": 2615 | |
| }, | |
| { | |
| "epoch": 0.7547441575744481, | |
| "grad_norm": 0.02742316326439345, | |
| "learning_rate": 3.453466313060322e-06, | |
| "loss": 0.0997, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.756184509020201, | |
| "grad_norm": 0.03242870927580512, | |
| "learning_rate": 3.4155278020654502e-06, | |
| "loss": 0.101, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 0.7576248604659537, | |
| "grad_norm": 0.02846968206424331, | |
| "learning_rate": 3.3777558683781795e-06, | |
| "loss": 0.1002, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 0.7590652119117065, | |
| "grad_norm": 0.025257296374579147, | |
| "learning_rate": 3.3401514675719815e-06, | |
| "loss": 0.0969, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 0.7605055633574592, | |
| "grad_norm": 0.02845603161574245, | |
| "learning_rate": 3.302715550982014e-06, | |
| "loss": 0.0998, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.7619459148032119, | |
| "grad_norm": 0.029606705311034962, | |
| "learning_rate": 3.2654490656810256e-06, | |
| "loss": 0.0965, | |
| "step": 2645 | |
| }, | |
| { | |
| "epoch": 0.7633862662489648, | |
| "grad_norm": 0.029342507148503363, | |
| "learning_rate": 3.228352954455406e-06, | |
| "loss": 0.0966, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.7648266176947175, | |
| "grad_norm": 0.02856949602254836, | |
| "learning_rate": 3.1914281557813386e-06, | |
| "loss": 0.0966, | |
| "step": 2655 | |
| }, | |
| { | |
| "epoch": 0.7662669691404703, | |
| "grad_norm": 0.03497921521636249, | |
| "learning_rate": 3.1546756038010507e-06, | |
| "loss": 0.0981, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.767707320586223, | |
| "grad_norm": 0.030940811674726062, | |
| "learning_rate": 3.1180962282991976e-06, | |
| "loss": 0.1008, | |
| "step": 2665 | |
| }, | |
| { | |
| "epoch": 0.7691476720319758, | |
| "grad_norm": 0.02732290490010739, | |
| "learning_rate": 3.081690954679313e-06, | |
| "loss": 0.0968, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.7705880234777286, | |
| "grad_norm": 0.029043196623878744, | |
| "learning_rate": 3.0454607039404206e-06, | |
| "loss": 0.0957, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 0.7720283749234813, | |
| "grad_norm": 0.029602576168767845, | |
| "learning_rate": 3.0094063926537233e-06, | |
| "loss": 0.0993, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.7734687263692341, | |
| "grad_norm": 0.02834425216635453, | |
| "learning_rate": 2.973528932939429e-06, | |
| "loss": 0.0954, | |
| "step": 2685 | |
| }, | |
| { | |
| "epoch": 0.7749090778149869, | |
| "grad_norm": 0.02696740012171304, | |
| "learning_rate": 2.937829232443654e-06, | |
| "loss": 0.0974, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 0.7763494292607396, | |
| "grad_norm": 0.029300105549739816, | |
| "learning_rate": 2.9023081943154753e-06, | |
| "loss": 0.1012, | |
| "step": 2695 | |
| }, | |
| { | |
| "epoch": 0.7777897807064924, | |
| "grad_norm": 0.025590035684474745, | |
| "learning_rate": 2.86696671718408e-06, | |
| "loss": 0.0954, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.7792301321522451, | |
| "grad_norm": 0.028340578845450066, | |
| "learning_rate": 2.8318056951360294e-06, | |
| "loss": 0.0978, | |
| "step": 2705 | |
| }, | |
| { | |
| "epoch": 0.780670483597998, | |
| "grad_norm": 0.027993276158026614, | |
| "learning_rate": 2.7968260176926407e-06, | |
| "loss": 0.1014, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 0.7821108350437507, | |
| "grad_norm": 0.03153867114181106, | |
| "learning_rate": 2.762028569787485e-06, | |
| "loss": 0.1051, | |
| "step": 2715 | |
| }, | |
| { | |
| "epoch": 0.7835511864895034, | |
| "grad_norm": 0.03024494353093925, | |
| "learning_rate": 2.7274142317439956e-06, | |
| "loss": 0.0979, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.7849915379352562, | |
| "grad_norm": 0.02889386857808468, | |
| "learning_rate": 2.6929838792532035e-06, | |
| "loss": 0.102, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 0.7864318893810089, | |
| "grad_norm": 0.028444248618397415, | |
| "learning_rate": 2.6587383833515746e-06, | |
| "loss": 0.097, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.7878722408267618, | |
| "grad_norm": 0.03018963515700107, | |
| "learning_rate": 2.6246786103989887e-06, | |
| "loss": 0.0969, | |
| "step": 2735 | |
| }, | |
| { | |
| "epoch": 0.7893125922725145, | |
| "grad_norm": 0.028448487698969887, | |
| "learning_rate": 2.590805422056807e-06, | |
| "loss": 0.0976, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.7907529437182672, | |
| "grad_norm": 0.025980334968211127, | |
| "learning_rate": 2.5571196752660733e-06, | |
| "loss": 0.098, | |
| "step": 2745 | |
| }, | |
| { | |
| "epoch": 0.79219329516402, | |
| "grad_norm": 0.0279120349990706, | |
| "learning_rate": 2.5236222222258455e-06, | |
| "loss": 0.1006, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.7936336466097728, | |
| "grad_norm": 0.03043213886693375, | |
| "learning_rate": 2.4903139103716365e-06, | |
| "loss": 0.0961, | |
| "step": 2755 | |
| }, | |
| { | |
| "epoch": 0.7950739980555256, | |
| "grad_norm": 0.027713327075462212, | |
| "learning_rate": 2.4571955823539617e-06, | |
| "loss": 0.0927, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.7965143495012783, | |
| "grad_norm": 0.03200709702140973, | |
| "learning_rate": 2.424268076017032e-06, | |
| "loss": 0.099, | |
| "step": 2765 | |
| }, | |
| { | |
| "epoch": 0.797954700947031, | |
| "grad_norm": 0.02862031576494855, | |
| "learning_rate": 2.3915322243775564e-06, | |
| "loss": 0.095, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 0.7993950523927839, | |
| "grad_norm": 0.030880108006987813, | |
| "learning_rate": 2.3589888556036623e-06, | |
| "loss": 0.1012, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 0.8008354038385366, | |
| "grad_norm": 0.028581095269047594, | |
| "learning_rate": 2.3266387929939525e-06, | |
| "loss": 0.1004, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.8022757552842894, | |
| "grad_norm": 0.029454825149171658, | |
| "learning_rate": 2.294482854956672e-06, | |
| "loss": 0.0984, | |
| "step": 2785 | |
| }, | |
| { | |
| "epoch": 0.8037161067300421, | |
| "grad_norm": 0.03006431233771053, | |
| "learning_rate": 2.2625218549890014e-06, | |
| "loss": 0.0995, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.8051564581757948, | |
| "grad_norm": 0.027799037780853072, | |
| "learning_rate": 2.230756601656481e-06, | |
| "loss": 0.1018, | |
| "step": 2795 | |
| }, | |
| { | |
| "epoch": 0.8065968096215477, | |
| "grad_norm": 0.03649536849392369, | |
| "learning_rate": 2.1991878985725566e-06, | |
| "loss": 0.0936, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.8080371610673004, | |
| "grad_norm": 0.028269913936191833, | |
| "learning_rate": 2.167816544378244e-06, | |
| "loss": 0.1009, | |
| "step": 2805 | |
| }, | |
| { | |
| "epoch": 0.8094775125130532, | |
| "grad_norm": 0.028604988396104106, | |
| "learning_rate": 2.1366433327219284e-06, | |
| "loss": 0.1025, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 0.8109178639588059, | |
| "grad_norm": 0.02876137453572099, | |
| "learning_rate": 2.105669052239274e-06, | |
| "loss": 0.1029, | |
| "step": 2815 | |
| }, | |
| { | |
| "epoch": 0.8123582154045587, | |
| "grad_norm": 0.028447843531367326, | |
| "learning_rate": 2.0748944865333033e-06, | |
| "loss": 0.0989, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.8137985668503115, | |
| "grad_norm": 0.02868959528432683, | |
| "learning_rate": 2.0443204141545393e-06, | |
| "loss": 0.0991, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 0.8152389182960642, | |
| "grad_norm": 0.030631865795776963, | |
| "learning_rate": 2.013947608581327e-06, | |
| "loss": 0.0988, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 0.816679269741817, | |
| "grad_norm": 0.028274634163983327, | |
| "learning_rate": 1.983776838200262e-06, | |
| "loss": 0.0967, | |
| "step": 2835 | |
| }, | |
| { | |
| "epoch": 0.8181196211875698, | |
| "grad_norm": 0.03127569269957067, | |
| "learning_rate": 1.9538088662867495e-06, | |
| "loss": 0.0989, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.8195599726333225, | |
| "grad_norm": 0.028435906385402127, | |
| "learning_rate": 1.924044450985706e-06, | |
| "loss": 0.1035, | |
| "step": 2845 | |
| }, | |
| { | |
| "epoch": 0.8210003240790753, | |
| "grad_norm": 0.030905808512586234, | |
| "learning_rate": 1.8944843452923546e-06, | |
| "loss": 0.1031, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.822440675524828, | |
| "grad_norm": 0.02821159641559338, | |
| "learning_rate": 1.8651292970332003e-06, | |
| "loss": 0.0975, | |
| "step": 2855 | |
| }, | |
| { | |
| "epoch": 0.8238810269705809, | |
| "grad_norm": 0.026301972641988965, | |
| "learning_rate": 1.835980048847098e-06, | |
| "loss": 0.1016, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.8253213784163336, | |
| "grad_norm": 0.02850407318953127, | |
| "learning_rate": 1.8070373381664752e-06, | |
| "loss": 0.0957, | |
| "step": 2865 | |
| }, | |
| { | |
| "epoch": 0.8267617298620864, | |
| "grad_norm": 0.03287640405240904, | |
| "learning_rate": 1.77830189719866e-06, | |
| "loss": 0.0974, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 0.8282020813078391, | |
| "grad_norm": 0.02979639575966765, | |
| "learning_rate": 1.7497744529073712e-06, | |
| "loss": 0.0962, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 0.8296424327535918, | |
| "grad_norm": 0.028883353854663875, | |
| "learning_rate": 1.721455726994321e-06, | |
| "loss": 0.1033, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.8310827841993447, | |
| "grad_norm": 0.027623128069878516, | |
| "learning_rate": 1.6933464358809593e-06, | |
| "loss": 0.0915, | |
| "step": 2885 | |
| }, | |
| { | |
| "epoch": 0.8325231356450974, | |
| "grad_norm": 0.031518406493856295, | |
| "learning_rate": 1.6654472906903486e-06, | |
| "loss": 0.0953, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 0.8339634870908502, | |
| "grad_norm": 0.025571212721863242, | |
| "learning_rate": 1.637758997229173e-06, | |
| "loss": 0.0939, | |
| "step": 2895 | |
| }, | |
| { | |
| "epoch": 0.8354038385366029, | |
| "grad_norm": 0.026694901604252565, | |
| "learning_rate": 1.6102822559698828e-06, | |
| "loss": 0.0947, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.8368441899823557, | |
| "grad_norm": 0.0277750741435232, | |
| "learning_rate": 1.5830177620329712e-06, | |
| "loss": 0.0989, | |
| "step": 2905 | |
| }, | |
| { | |
| "epoch": 0.8382845414281085, | |
| "grad_norm": 0.02924974903749895, | |
| "learning_rate": 1.5559662051694002e-06, | |
| "loss": 0.0986, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.8397248928738612, | |
| "grad_norm": 0.032423419131996915, | |
| "learning_rate": 1.5291282697431353e-06, | |
| "loss": 0.0989, | |
| "step": 2915 | |
| }, | |
| { | |
| "epoch": 0.841165244319614, | |
| "grad_norm": 0.030642374976224046, | |
| "learning_rate": 1.502504634713835e-06, | |
| "loss": 0.0963, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.8426055957653668, | |
| "grad_norm": 0.02981716833954207, | |
| "learning_rate": 1.4760959736196834e-06, | |
| "loss": 0.0961, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 0.8440459472111195, | |
| "grad_norm": 0.028126597558143722, | |
| "learning_rate": 1.4499029545603472e-06, | |
| "loss": 0.0996, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 0.8454862986568723, | |
| "grad_norm": 0.030061407833917978, | |
| "learning_rate": 1.423926240180068e-06, | |
| "loss": 0.1027, | |
| "step": 2935 | |
| }, | |
| { | |
| "epoch": 0.846926650102625, | |
| "grad_norm": 0.030976625673774977, | |
| "learning_rate": 1.3981664876509028e-06, | |
| "loss": 0.0995, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.8483670015483779, | |
| "grad_norm": 0.030419140568639338, | |
| "learning_rate": 1.3726243486560975e-06, | |
| "loss": 0.1049, | |
| "step": 2945 | |
| }, | |
| { | |
| "epoch": 0.8498073529941306, | |
| "grad_norm": 0.02669531962544557, | |
| "learning_rate": 1.3473004693736037e-06, | |
| "loss": 0.1026, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.8512477044398833, | |
| "grad_norm": 0.028011138842770462, | |
| "learning_rate": 1.3221954904597256e-06, | |
| "loss": 0.0957, | |
| "step": 2955 | |
| }, | |
| { | |
| "epoch": 0.8526880558856361, | |
| "grad_norm": 0.029219102403077252, | |
| "learning_rate": 1.2973100470329159e-06, | |
| "loss": 0.0989, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 0.8541284073313888, | |
| "grad_norm": 0.028252079128452696, | |
| "learning_rate": 1.272644768657707e-06, | |
| "loss": 0.096, | |
| "step": 2965 | |
| }, | |
| { | |
| "epoch": 0.8555687587771417, | |
| "grad_norm": 0.028035615158873324, | |
| "learning_rate": 1.248200279328784e-06, | |
| "loss": 0.0985, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 0.8570091102228944, | |
| "grad_norm": 0.03053162765030892, | |
| "learning_rate": 1.223977197455204e-06, | |
| "loss": 0.1006, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 0.8584494616686471, | |
| "grad_norm": 0.028648955173488484, | |
| "learning_rate": 1.1999761358447403e-06, | |
| "loss": 0.0994, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 0.8598898131143999, | |
| "grad_norm": 0.02748195928805478, | |
| "learning_rate": 1.1761977016883897e-06, | |
| "loss": 0.0958, | |
| "step": 2985 | |
| }, | |
| { | |
| "epoch": 0.8613301645601527, | |
| "grad_norm": 0.028621187530613948, | |
| "learning_rate": 1.152642496544998e-06, | |
| "loss": 0.0958, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 0.8627705160059055, | |
| "grad_norm": 0.027853067803610992, | |
| "learning_rate": 1.1293111163260639e-06, | |
| "loss": 0.0997, | |
| "step": 2995 | |
| }, | |
| { | |
| "epoch": 0.8642108674516582, | |
| "grad_norm": 0.027187976560834638, | |
| "learning_rate": 1.1062041512806409e-06, | |
| "loss": 0.1028, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.8642108674516582, | |
| "eval_loss": 0.09996497631072998, | |
| "eval_runtime": 862.256, | |
| "eval_samples_per_second": 2.092, | |
| "eval_steps_per_second": 0.523, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.8656512188974109, | |
| "grad_norm": 0.026511113426348355, | |
| "learning_rate": 1.0833221859804188e-06, | |
| "loss": 0.0976, | |
| "step": 3005 | |
| }, | |
| { | |
| "epoch": 0.8670915703431638, | |
| "grad_norm": 0.02987993551142259, | |
| "learning_rate": 1.0606657993049253e-06, | |
| "loss": 0.0966, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 0.8685319217889165, | |
| "grad_norm": 0.029689732788246524, | |
| "learning_rate": 1.0382355644268871e-06, | |
| "loss": 0.1041, | |
| "step": 3015 | |
| }, | |
| { | |
| "epoch": 0.8699722732346693, | |
| "grad_norm": 0.0267869386208169, | |
| "learning_rate": 1.0160320487977349e-06, | |
| "loss": 0.0966, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 0.871412624680422, | |
| "grad_norm": 0.032293109106566735, | |
| "learning_rate": 9.940558141332323e-07, | |
| "loss": 0.1048, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 0.8728529761261747, | |
| "grad_norm": 0.026383256040976497, | |
| "learning_rate": 9.723074163992774e-07, | |
| "loss": 0.0988, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 0.8742933275719276, | |
| "grad_norm": 0.028451018898224757, | |
| "learning_rate": 9.507874057978339e-07, | |
| "loss": 0.0974, | |
| "step": 3035 | |
| }, | |
| { | |
| "epoch": 0.8757336790176803, | |
| "grad_norm": 0.03180099043766917, | |
| "learning_rate": 9.294963267530177e-07, | |
| "loss": 0.097, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 0.8771740304634331, | |
| "grad_norm": 0.030029444501747108, | |
| "learning_rate": 9.084347178973107e-07, | |
| "loss": 0.0963, | |
| "step": 3045 | |
| }, | |
| { | |
| "epoch": 0.8786143819091858, | |
| "grad_norm": 0.030459369725609208, | |
| "learning_rate": 8.876031120579454e-07, | |
| "loss": 0.0985, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.8800547333549386, | |
| "grad_norm": 0.030593158083603012, | |
| "learning_rate": 8.670020362434229e-07, | |
| "loss": 0.0975, | |
| "step": 3055 | |
| }, | |
| { | |
| "epoch": 0.8814950848006914, | |
| "grad_norm": 0.02863412252326483, | |
| "learning_rate": 8.466320116301752e-07, | |
| "loss": 0.0959, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 0.8829354362464441, | |
| "grad_norm": 0.028283718935748024, | |
| "learning_rate": 8.264935535493879e-07, | |
| "loss": 0.0956, | |
| "step": 3065 | |
| }, | |
| { | |
| "epoch": 0.8843757876921969, | |
| "grad_norm": 0.032405657391703896, | |
| "learning_rate": 8.065871714739581e-07, | |
| "loss": 0.1016, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 0.8858161391379497, | |
| "grad_norm": 0.02841639634176925, | |
| "learning_rate": 7.869133690056063e-07, | |
| "loss": 0.0982, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 0.8872564905837024, | |
| "grad_norm": 0.026380393134296947, | |
| "learning_rate": 7.67472643862136e-07, | |
| "loss": 0.0949, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 0.8886968420294552, | |
| "grad_norm": 0.03051349318569555, | |
| "learning_rate": 7.482654878648465e-07, | |
| "loss": 0.1076, | |
| "step": 3085 | |
| }, | |
| { | |
| "epoch": 0.8901371934752079, | |
| "grad_norm": 0.026798564091589188, | |
| "learning_rate": 7.292923869260837e-07, | |
| "loss": 0.1009, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 0.8915775449209608, | |
| "grad_norm": 0.026243273967039876, | |
| "learning_rate": 7.105538210369467e-07, | |
| "loss": 0.0937, | |
| "step": 3095 | |
| }, | |
| { | |
| "epoch": 0.8930178963667135, | |
| "grad_norm": 0.027316057706649032, | |
| "learning_rate": 6.920502642551519e-07, | |
| "loss": 0.0981, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.8944582478124662, | |
| "grad_norm": 0.030461472244114565, | |
| "learning_rate": 6.737821846930403e-07, | |
| "loss": 0.1013, | |
| "step": 3105 | |
| }, | |
| { | |
| "epoch": 0.895898599258219, | |
| "grad_norm": 0.029378550013335232, | |
| "learning_rate": 6.557500445057252e-07, | |
| "loss": 0.1032, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 0.8973389507039717, | |
| "grad_norm": 0.03205510198240369, | |
| "learning_rate": 6.379542998794086e-07, | |
| "loss": 0.0942, | |
| "step": 3115 | |
| }, | |
| { | |
| "epoch": 0.8987793021497246, | |
| "grad_norm": 0.02919082875531507, | |
| "learning_rate": 6.203954010198387e-07, | |
| "loss": 0.1016, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 0.9002196535954773, | |
| "grad_norm": 0.028309778648473748, | |
| "learning_rate": 6.030737921409169e-07, | |
| "loss": 0.0972, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 0.90166000504123, | |
| "grad_norm": 0.02666645822561416, | |
| "learning_rate": 5.859899114534662e-07, | |
| "loss": 0.0958, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 0.9031003564869828, | |
| "grad_norm": 0.02914798086446344, | |
| "learning_rate": 5.691441911541385e-07, | |
| "loss": 0.0989, | |
| "step": 3135 | |
| }, | |
| { | |
| "epoch": 0.9045407079327356, | |
| "grad_norm": 0.029321419229845342, | |
| "learning_rate": 5.525370574144873e-07, | |
| "loss": 0.0943, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 0.9059810593784884, | |
| "grad_norm": 0.029224751094801393, | |
| "learning_rate": 5.361689303701767e-07, | |
| "loss": 0.0927, | |
| "step": 3145 | |
| }, | |
| { | |
| "epoch": 0.9074214108242411, | |
| "grad_norm": 0.028920548399641463, | |
| "learning_rate": 5.200402241103674e-07, | |
| "loss": 0.0973, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.9088617622699939, | |
| "grad_norm": 0.026508682335799016, | |
| "learning_rate": 5.041513466672254e-07, | |
| "loss": 0.0911, | |
| "step": 3155 | |
| }, | |
| { | |
| "epoch": 0.9103021137157467, | |
| "grad_norm": 0.0286913163547185, | |
| "learning_rate": 4.885027000056075e-07, | |
| "loss": 0.1003, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 0.9117424651614994, | |
| "grad_norm": 0.028408650516893636, | |
| "learning_rate": 4.730946800128888e-07, | |
| "loss": 0.0991, | |
| "step": 3165 | |
| }, | |
| { | |
| "epoch": 0.9131828166072522, | |
| "grad_norm": 0.027915924830151973, | |
| "learning_rate": 4.5792767648895396e-07, | |
| "loss": 0.1008, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 0.9146231680530049, | |
| "grad_norm": 0.03016832518348953, | |
| "learning_rate": 4.4300207313632713e-07, | |
| "loss": 0.0979, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 0.9160635194987578, | |
| "grad_norm": 0.025345673276194875, | |
| "learning_rate": 4.2831824755046994e-07, | |
| "loss": 0.0963, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 0.9175038709445105, | |
| "grad_norm": 0.026699911404801765, | |
| "learning_rate": 4.138765712102299e-07, | |
| "loss": 0.0949, | |
| "step": 3185 | |
| }, | |
| { | |
| "epoch": 0.9189442223902632, | |
| "grad_norm": 0.028093887453801158, | |
| "learning_rate": 3.9967740946843523e-07, | |
| "loss": 0.0962, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 0.920384573836016, | |
| "grad_norm": 0.030027557409672574, | |
| "learning_rate": 3.8572112154266593e-07, | |
| "loss": 0.1003, | |
| "step": 3195 | |
| }, | |
| { | |
| "epoch": 0.9218249252817687, | |
| "grad_norm": 0.030720426408657343, | |
| "learning_rate": 3.7200806050614714e-07, | |
| "loss": 0.0966, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.9232652767275216, | |
| "grad_norm": 0.028926025323559437, | |
| "learning_rate": 3.585385732788327e-07, | |
| "loss": 0.0957, | |
| "step": 3205 | |
| }, | |
| { | |
| "epoch": 0.9247056281732743, | |
| "grad_norm": 0.03702215777030734, | |
| "learning_rate": 3.453130006186234e-07, | |
| "loss": 0.0982, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 0.926145979619027, | |
| "grad_norm": 0.029009798104548613, | |
| "learning_rate": 3.3233167711274496e-07, | |
| "loss": 0.1018, | |
| "step": 3215 | |
| }, | |
| { | |
| "epoch": 0.9275863310647798, | |
| "grad_norm": 0.026198880889564888, | |
| "learning_rate": 3.1959493116928473e-07, | |
| "loss": 0.0987, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 0.9290266825105326, | |
| "grad_norm": 0.028949046641791966, | |
| "learning_rate": 3.0710308500888184e-07, | |
| "loss": 0.1009, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 0.9304670339562854, | |
| "grad_norm": 0.03211145771064907, | |
| "learning_rate": 2.948564546565791e-07, | |
| "loss": 0.0964, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 0.9319073854020381, | |
| "grad_norm": 0.031242518818499414, | |
| "learning_rate": 2.828553499338227e-07, | |
| "loss": 0.0946, | |
| "step": 3235 | |
| }, | |
| { | |
| "epoch": 0.9333477368477908, | |
| "grad_norm": 0.031264506938530154, | |
| "learning_rate": 2.71100074450632e-07, | |
| "loss": 0.0958, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 0.9347880882935437, | |
| "grad_norm": 0.029366673066917617, | |
| "learning_rate": 2.595909255979079e-07, | |
| "loss": 0.0981, | |
| "step": 3245 | |
| }, | |
| { | |
| "epoch": 0.9362284397392964, | |
| "grad_norm": 0.0265988622897014, | |
| "learning_rate": 2.4832819453992073e-07, | |
| "loss": 0.099, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.9376687911850492, | |
| "grad_norm": 0.027150460711859906, | |
| "learning_rate": 2.3731216620693554e-07, | |
| "loss": 0.1026, | |
| "step": 3255 | |
| }, | |
| { | |
| "epoch": 0.9391091426308019, | |
| "grad_norm": 0.03127106760239122, | |
| "learning_rate": 2.2654311928800965e-07, | |
| "loss": 0.093, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 0.9405494940765546, | |
| "grad_norm": 0.030758685060327442, | |
| "learning_rate": 2.1602132622393745e-07, | |
| "loss": 0.1015, | |
| "step": 3265 | |
| }, | |
| { | |
| "epoch": 0.9419898455223075, | |
| "grad_norm": 0.028184366940994177, | |
| "learning_rate": 2.0574705320036025e-07, | |
| "loss": 0.1074, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 0.9434301969680602, | |
| "grad_norm": 0.02849030323321046, | |
| "learning_rate": 1.9572056014103281e-07, | |
| "loss": 0.1001, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 0.944870548413813, | |
| "grad_norm": 0.03266549303159609, | |
| "learning_rate": 1.8594210070124852e-07, | |
| "loss": 0.1045, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 0.9463108998595657, | |
| "grad_norm": 0.026988212769410665, | |
| "learning_rate": 1.7641192226141913e-07, | |
| "loss": 0.0937, | |
| "step": 3285 | |
| }, | |
| { | |
| "epoch": 0.9477512513053185, | |
| "grad_norm": 0.02529946877891728, | |
| "learning_rate": 1.671302659208185e-07, | |
| "loss": 0.098, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 0.9491916027510713, | |
| "grad_norm": 0.03314606321848586, | |
| "learning_rate": 1.58097366491482e-07, | |
| "loss": 0.1009, | |
| "step": 3295 | |
| }, | |
| { | |
| "epoch": 0.950631954196824, | |
| "grad_norm": 0.027540135720360067, | |
| "learning_rate": 1.4931345249226792e-07, | |
| "loss": 0.0967, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.9520723056425768, | |
| "grad_norm": 0.028602952429147437, | |
| "learning_rate": 1.407787461430743e-07, | |
| "loss": 0.1042, | |
| "step": 3305 | |
| }, | |
| { | |
| "epoch": 0.9535126570883296, | |
| "grad_norm": 0.028879029680259325, | |
| "learning_rate": 1.324934633592201e-07, | |
| "loss": 0.0945, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 0.9549530085340823, | |
| "grad_norm": 0.02860475239355599, | |
| "learning_rate": 1.2445781374597842e-07, | |
| "loss": 0.1017, | |
| "step": 3315 | |
| }, | |
| { | |
| "epoch": 0.9563933599798351, | |
| "grad_norm": 0.026737800323851892, | |
| "learning_rate": 1.1667200059327644e-07, | |
| "loss": 0.0971, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 0.9578337114255878, | |
| "grad_norm": 0.029364181788732343, | |
| "learning_rate": 1.0913622087055264e-07, | |
| "loss": 0.1005, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 0.9592740628713406, | |
| "grad_norm": 0.03131909258187323, | |
| "learning_rate": 1.0185066522177545e-07, | |
| "loss": 0.092, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 0.9607144143170934, | |
| "grad_norm": 0.029983775954717285, | |
| "learning_rate": 9.481551796061472e-08, | |
| "loss": 0.0989, | |
| "step": 3335 | |
| }, | |
| { | |
| "epoch": 0.9621547657628461, | |
| "grad_norm": 0.030901618242872846, | |
| "learning_rate": 8.803095706578335e-08, | |
| "loss": 0.0987, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 0.9635951172085989, | |
| "grad_norm": 0.027894545371751872, | |
| "learning_rate": 8.149715417653414e-08, | |
| "loss": 0.0969, | |
| "step": 3345 | |
| }, | |
| { | |
| "epoch": 0.9650354686543516, | |
| "grad_norm": 0.027564458728654692, | |
| "learning_rate": 7.521427458831776e-08, | |
| "loss": 0.099, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.9664758201001045, | |
| "grad_norm": 0.025676377662014388, | |
| "learning_rate": 6.918247724859939e-08, | |
| "loss": 0.0941, | |
| "step": 3355 | |
| }, | |
| { | |
| "epoch": 0.9679161715458572, | |
| "grad_norm": 0.030832764931546873, | |
| "learning_rate": 6.340191475283753e-08, | |
| "loss": 0.0988, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 0.9693565229916099, | |
| "grad_norm": 0.03090619704744881, | |
| "learning_rate": 5.787273334062593e-08, | |
| "loss": 0.1, | |
| "step": 3365 | |
| }, | |
| { | |
| "epoch": 0.9707968744373627, | |
| "grad_norm": 0.028989234177967914, | |
| "learning_rate": 5.259507289199328e-08, | |
| "loss": 0.1042, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 0.9722372258831155, | |
| "grad_norm": 0.031096318018671423, | |
| "learning_rate": 4.756906692386043e-08, | |
| "loss": 0.1015, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 0.9736775773288683, | |
| "grad_norm": 0.031160065189087587, | |
| "learning_rate": 4.2794842586670884e-08, | |
| "loss": 0.0987, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 0.975117928774621, | |
| "grad_norm": 0.027856429944083137, | |
| "learning_rate": 3.827252066116338e-08, | |
| "loss": 0.0934, | |
| "step": 3385 | |
| }, | |
| { | |
| "epoch": 0.9765582802203737, | |
| "grad_norm": 0.0292530610363585, | |
| "learning_rate": 3.400221555532768e-08, | |
| "loss": 0.0931, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 0.9779986316661265, | |
| "grad_norm": 0.030499523184600807, | |
| "learning_rate": 2.998403530150018e-08, | |
| "loss": 0.1011, | |
| "step": 3395 | |
| }, | |
| { | |
| "epoch": 0.9794389831118793, | |
| "grad_norm": 0.030425018653641535, | |
| "learning_rate": 2.6218081553638363e-08, | |
| "loss": 0.1015, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.9808793345576321, | |
| "grad_norm": 0.028935900226943704, | |
| "learning_rate": 2.2704449584745046e-08, | |
| "loss": 0.1015, | |
| "step": 3405 | |
| }, | |
| { | |
| "epoch": 0.9823196860033848, | |
| "grad_norm": 0.031000647192898592, | |
| "learning_rate": 1.9443228284455882e-08, | |
| "loss": 0.0967, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 0.9837600374491375, | |
| "grad_norm": 0.02615522374107861, | |
| "learning_rate": 1.6434500156800037e-08, | |
| "loss": 0.098, | |
| "step": 3415 | |
| }, | |
| { | |
| "epoch": 0.9852003888948904, | |
| "grad_norm": 0.0296512129505122, | |
| "learning_rate": 1.3678341318100751e-08, | |
| "loss": 0.0941, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 0.9866407403406431, | |
| "grad_norm": 0.027814116523255324, | |
| "learning_rate": 1.1174821495059106e-08, | |
| "loss": 0.0959, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 0.9880810917863959, | |
| "grad_norm": 0.03492776644651804, | |
| "learning_rate": 8.924004022986543e-09, | |
| "loss": 0.0953, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 0.9895214432321486, | |
| "grad_norm": 0.02820257683900699, | |
| "learning_rate": 6.9259458442005875e-09, | |
| "loss": 0.1026, | |
| "step": 3435 | |
| }, | |
| { | |
| "epoch": 0.9909617946779014, | |
| "grad_norm": 0.030304015663341097, | |
| "learning_rate": 5.180697506587118e-09, | |
| "loss": 0.0955, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 0.9924021461236542, | |
| "grad_norm": 0.026993412863199263, | |
| "learning_rate": 3.688303162322493e-09, | |
| "loss": 0.0969, | |
| "step": 3445 | |
| }, | |
| { | |
| "epoch": 0.9938424975694069, | |
| "grad_norm": 0.02671367368351453, | |
| "learning_rate": 2.44880056675334e-09, | |
| "loss": 0.0921, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.9952828490151597, | |
| "grad_norm": 0.027626479574708512, | |
| "learning_rate": 1.4622210774428714e-09, | |
| "loss": 0.0986, | |
| "step": 3455 | |
| }, | |
| { | |
| "epoch": 0.9967232004609125, | |
| "grad_norm": 0.02933812040981167, | |
| "learning_rate": 7.285896533770765e-10, | |
| "loss": 0.0992, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 0.9981635519066653, | |
| "grad_norm": 0.027258112232128342, | |
| "learning_rate": 2.479248543363344e-10, | |
| "loss": 0.0967, | |
| "step": 3465 | |
| }, | |
| { | |
| "epoch": 0.999603903352418, | |
| "grad_norm": 0.029806576902039788, | |
| "learning_rate": 2.0238840421349382e-11, | |
| "loss": 0.0979, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 0.9998919736415686, | |
| "step": 3471, | |
| "total_flos": 2.660820529446912e+16, | |
| "train_loss": 0.2432632086317191, | |
| "train_runtime": 212607.6799, | |
| "train_samples_per_second": 1.045, | |
| "train_steps_per_second": 0.016 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 3471, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.660820529446912e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |